Long-Term Safety, Clinical and Immunological Outcomes in Primary Membranous Nephropathy with Severe Renal Impairment Treated with Cyclophosphamide and Steroid-Based Regimen

Membranous nephropathy (MN) is considered a rare disease. Although up to 40% with primary MN (PMN) may develop spontaneous remission [1], PMN remains a leading cause of nephrotic syndrome and progressive kidney disease. Development of end-stage kidney disease (ESKD) is the most frequent and disabling complication of PMN; with approximately 15–40% estimated to progress to (ESKD) over 5–15 years [2]. Over the last 20 years, incidence of ESKD needing renal replacement therapy (RRT) has not changed in the UK (UK Renal Registry, personal correspondence). Immunosuppression (ImS) can prevent loss of kidney function and progression to kidney failure but is associated with side effects which require monitoring and costs. Therefore, in clinical practice, a combination of risk factors is considered to assess risk of progression and eligibility for starting ImS therapy with these agents typically used in patients with high or very high-risk factors [3].

Accumulating evidence suggests that patients with high levels of proteinuria, deteriorating kidney function and higher anti-PLA₂R-Ab levels are at higher risk of developing kidney failure and are considered for immunosuppressive therapy. Conventional approaches to ImS in PMN include calcineurin inhibitors, combined cyclophosphamide and steroid therapy and rituximab, but outcomes of treatment have not been examined systematically in patients with eGFR ≤30 mL/min/1.73 m² until recently [4‒6]. Although KDIGO recommendations suggest adding ImS to supportive care for those with rapid deteriorating kidney function, current treatment algorithms avoid ImS in patients with eGFR less than 30 mL/min citing potential risks and claiming reduced chances of clinical response [3, 7]. This recommendation is based on paucity of evidence on effectiveness or toxicity of such regimes in PMN. A plethora of recent interventional drug trials such as MENTOR [8], GEMRITUX [9], STARMEN [10], RI-CYCLO [11], NCT04154787 [12], and NCT04733040 [13] studies have universally excluded patients with eGFR <30 mL/min/1.73 m², further limiting evidence of benefit in this group of patients. It is well established that a 6-month treatment with a combination of cyclophosphamide and steroids (Ponticelli regime) can induce remission of disease activity in 70–90% of patients and protect decline of renal function long term [14, 15]. Variations of the Ponticelli regime are in routine clinical use across different centers [16, 17]. Although published RCTs on this regimen report good tolerance and infrequent side effects, its use is restricted to high-risk patients due to perceived toxicity profile. Patients with advanced kidney disease, despite being at high risk of progression have little choice therefore but to plan for future RRT. In this study, we analyzed the efficacy and long-term outcomes of combined steroids and cyclophosphamide administered to consecutive patients with a confirmed diagnosis of PMN and eGFR ≤30 mL/min.

18 consecutive patients were included in the study with (a) biopsy-proven PMN during 2014 and 2019, (b) developed advanced renal dysfunction (eGFR of ≤30 mL/min/1.73 m²), and (c) initiated therapy with combination of cyclophosphamide and steroids. All patients were followed up at Manchester Royal infirmary (Greater Manchester East Sector Renal Network). Screening for systemic lupus erythematosus and hepatitis was routinely undertaken, while investigations for malignancy were undertaken if symptoms were noted on a detailed clinical encounter at any point. Renal vein thrombosis was excluded with screening imaging (online suppl. Fig. 1; for all online suppl. material, see www.karger.com/doi/10.1159/000529605). All patients received standard maximal tolerated RAS inhibition. Anti-PLA₂R-Ab levels were measured at the time of ImS and 1-year posttreatment in 12 and 16 patients, respectively, from 2013 as the test became routinely available in clinical practice [18]. The study size was pragmatic based on all eligible patients.

Immunosuppressive regimen consisted of up to 6 doses of intravenous cyclophosphamide doses of 10 mg/kg (if aged under 70 years) and 7.5 mg/kg (if aged over 70 years) given at monthly intervals and oral prednisolone therapy for 6 months (at a dose of 0.75 mg/kg daily, up to 60 mg/day – with gradual tapering to 0.5 mg/kg/day by 3 months and 0.1 mg/kg/day by 6 months). The treatment course with cyclophosphamide was extended to a maximum of 9 doses or reduced to three doses based on the patient’s response and occurrence of adverse events. Patients received prophylaxis with co-Trimoxazole and fluconazole for the duration of cyclophosphamide therapy and gastric prophylaxis was undertaken with proton pump inhibitors. As part of standard clinical care, all patients received clinic follow-up with monitoring for toxicity and effectiveness every 2-weeks during pulse cyclophosphamide therapy and at least every 3 months thereafter. We have excluded patients who were put on other ImS treatment given the variability in immunosuppressive regimes used.

Clinical and immunological data were recorded at time points including

• a.

  at the time of presentation

• b.

  at the time of initiation of ImS

• c.

  1-year posttreatment, and

• d.

  at last follow-up.

Outcomes including patient survival, initiation of RRT, hospital admissions, infections, and malignancies were recorded. Glomerular filtration rate was estimated using the CKD-EPI formula. Serum anti-PLA₂R-Ab levels were measured using enzyme-linked immunosorbent assay, using 14 RU/mL positivity threshold.

Outcomes post-cyclophosphamide were defined as

• Partial remission (PR) – urinary protein-creatinine ratio between 0.3 and 3 g/g with a decrease by at least 50% from the initial value.

• Complete remission (CR) – urinary protein level to less than 0.3 g/g.

• Relapse was defined as the reappearance of nephrotic range proteinuria.

• In addition, immunological remission was defined by the absence of circulating anti-PLA₂R-Abs.

Written and informed consent was obtained from all patients as per standard of clinical care. All patients were recruited into the studies exploring the immunological mechanisms in PMN, under ethics references 06/Q1401/5 and 10/H1008/10y. The study is reported in line with the STROBE guidelines.

Continuous data were summarized as mean (+/− St Dev) for normally distributed variables or median (with IQR). Categorical data were expressed in percentages and survival outcomes were described using Kaplan-Meier curves. Analyses were performed in the R statistical environment (Ref: R Core team [2014]. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-projgect.org/ [19] ).

Baseline demographic, biochemical, immunological parameters are outlined in Table 1. Median age was 68 (IQR 58–73) years, M:F ratio was 5:1; 16 patients were Caucasians. Median eGFR at presentation was 34 (IQR 27–45) mL/min/1.73 m² at presentation and was 23 (IQR 18–27) mL/min/1.73 m² at initiation of ImS. Median serum albumin at presentation was 2.2 (IQR 1.8–2.5) g/dL and at initiation of therapy was 2.2 (IQR 1.8–2.7) g/dL and urine protein-creatinine ratios were 8.4 (IQR 6.9–10.7) g/g at presentation and 8.8 (IQR 7.3–11.4) g/g at initiation of therapy. All patients received RAS inhibition, and ImS was initiated for disease progression.

Anti-PLA₂R-Ab levels were recorded in 12 patients, being positive in all with median of 301.50 (IQR 196.75–2,658) RU/mL at presentation. All 18 biopsies were adequately sampled with at least 10 glomeruli each. 10 biopsies showed a mild degree of chronicity, 2 biopsies (patients 14 and 15) showed moderate scarring, and 2 biopsies (patients 1 and 2) showed a severe degree of IFTA. Histological features of chronicity were graded by the degree of background interstitial fibrosis and tubular atrophy as mild (1), moderate (2), and severe (3). All patients were naïve to immune suppression before combination therapy except one subject who received cyclosporine for 3 months before using combined cyclophosphamide and steroids (patient 4 in Table 2).

Clinical parameters of disease activity during follow-up are outlined in Table 2 for individual patients and are shown cumulatively in (Fig. 1). PR, CR, and progression to RRT during follow-up with time are depicted in (Fig. 2).

During a median follow-up of 67 (IQR 27–80) months from combined immunosuppressive therapy, PR was achieved in 16 patients (89%) out of which 7 achieved CR (44%). 4 of the 16 (25%) patients who achieved PR and none of the 7 patients who achieved CR, suffered one or more relapses. 2 of the 18 patients (11%) (patients 1 and 3 in Table 2) had progressive renal impairment needing RRT.

eGFR increased by (30%) from median of 23 to 33 (IQR 26–46) mL/min, urine protein-creatinine ratio improved by (85%) from 8.8 (IQR 7.3–11.4) g/g to 1.3 (IQR 0.7–2.2) g/g, and serum albumin by (50%) from 2.2 (IQR 1.8–2.7) to 3.3 (2.6–4.6) g/dL during last follow-up. Patients initiated immunosuppressive therapy at a median of 3.5 (IQR 2.3–6.8) months after biopsy diagnosis.

Where achieved, time to partial and CR was 6 and 29 months, respectively (Fig. 2). One of the patients (patient 2) achieved CR and eGFR improved by 4 mL/min even though his biopsy on presentation showed a severe degree of interstitial fibrosis/tubular atrophy.

Ten (83%) out of the 12 patients, who had their anti-PLA2R-Ab levels measured at the time of ImS initiation, achieved an immunological remission (Fig. 3). One of the 2 patients who did not achieve immunological response with the treatment subsequently received rituximab therapy and achieved remission later. Another patient needed to initiate dialysis following which he developed a very early onset recurrence of MN posttransplant. This was treated with rituximab following which he achieved CR and maintained for 5-year follow-up.

Nine of 12 patients had a high anti-PLA₂R-Ab level on presentation (defined as >150 RU/mL). 6 of these patients (patients 6, 7, 10, 12, 17, and 18) achieved immunological remission at 1 year. All these 6 patients achieved PR. Interestingly, the time to both partial and CR did not seem to be affected by the initial anti-PLA₂R-Ab levels.

Five episodes of infections including 2 requiring hospital admissions were recorded including cellulitis, respiratory tract infections, and urosepsis (associated with a long-term catheter). All events were successfully treated, and patients were discharged home. During the long-term follow-up period of 57 months, 4 cancers were noted involving bladder, mesothelioma, sigmoid colon, and skin. 4 patients died (22.2%) (patients 1, 4, 6, and 7) at the time of last follow-up, with causes of death being cancer in 1, cerebrovascular accident in 1 (following PR) and 2 with ischemic heart disease following progression of renal dysfunction with RRT.

To our knowledge, this study is the first to systemically evaluate the effect of alkylating agent-based ImS for patients with PMN and an eGFR of ≤30 mL/min/1.73 m². In this study, we examined the outcomes, efficacy, and long-term safety of combined cyclophosphamide and steroid therapy. To limit the cumulative doses of both steroids and cyclophosphamide, we used intravenous pulse therapy with cyclophosphamide along with tapering doses of daily oral steroids, as previously described [16, 17]. Despite progressive renal disease with median eGFR of 23 mL/min at initiation of therapy, 16 of 18 patients (89%) had favorable renal outcomes following treatment with cyclophosphamide and prednisolone during a relatively long follow-up period of 5 years.

Published data have shown that 5–15% of patients on conservative management will progress to stage 5 chronic kidney disease within 5–10 years duration [2] in a cohort of patients with average eGFR of 75 mL/min/1.73 m². Despite markedly reduced GFR in the study cohort, only 11% progressed further requiring RRT. Balancing the risk-to-benefit profile in treating patients with PMN using ImS treatment has been at the heart of most recommendations to achieve the best outcomes. KDIGO recommendations suggest initiation of antiproteinuric treatment for 6 months for mild to moderate cases with possible spontaneous remission over time and to limit the exposure of immunosuppressive treatment to only patients presenting with life-threatening complications or a rapid decline in kidney function [3]. This approach assists in limiting exposure to potential toxicities that could be associated with immunosuppressive therapy. However, it is not known if such an approach in routine clinical management may result in loss of kidney function in patients with “moderate” and “severe” disease profile, or in those with comorbidities where clinicians and patients may be inclined to adopt a more cautious approach to immunosuppressive therapy.

Supportive care with observation alone may not be sufficient in patients presenting with severe renal dysfunction. Patients in this cohort with eGFR of 34 mL/min at presentation lost 11 mL/min prior to initiation of ImS (3 months of diagnosis). Among these 18 patients, a quarter of them waited for 6 months prior to initiation of ImS.

Use of ImS treatment in advanced kidney disease has not been thoroughly examined. The combination of cyclophosphamide and steroids is effective in achieving remission of nephrotic syndrome and stabilization of renal function with 72% and 92% remission rates in the two landmark trials. Ponticelli et al. [14] excluded patients with serum creatinine over 135 μmol/L in their study and mean plasma creatinine was 92 μmol/L; patients in Jha et al.'s [15] study had an MDRD eGFR of around 85 mL/min. Average proteinuria in the two studies was 6–7 g per 24 h.

Data from long-term trials using different ImS regimens for patients with an eGFR of >30 mL/min/1.73 m² showed a trend toward preserved renal function and achieving a state of remission over years of follow-up [15, 20]. Randomized controlled trials of immunosuppressive therapy in progressive MN showed the Ponticelli et al. [14] regimen using steroid and chlorambucil were superior to both placebo and cyclosporine [21]. There were more adverse events observed using the Ponticelli et al. [14] regimen attributed possibly to the use of chlorambucil rather than cyclophosphamide. But authors concluded that for a subset of patients with idiopathic MN and declining renal function, cyclosporine should be best avoided and 6 months of prednisolone and an alkylating agent is the best approach. In another report of a series of 15 patients with MN, authors showed that 2 patients presenting with an eGFR of <30 mL/min/1.73 m² at treatment achieved remission after 1 year of treatment using rituximab in combination with low dose cyclophosphamide and prednisolone, with no apparent adverse event [4]. The same group of researchers followed up a cohort of 60 patients with PLA₂R associated MN over a median of 3 years after being treated with the same ImS therapy combination. They found that all patients have achieved PR with an acceptable safety profile including 13% of patients who presented initially with an eGFR <30 mL/min/1.73 m²[5]. Hanset et al. [6], examined the efficacy of rituximab in 13 patients with an eGFR of <30 mL/min/1.73 m² at treatment. Over a median follow-up period of 17.8 months, eGFR improved from 18 mL/min at treatment to 23 mL/min. Nine (69%) patients responded and maintained remission with treatment. Four patients (31%) had progression to ESKD within 1 year (median of 6.7 months from treatment). Three of these patients had eGFR of <15 mL/min at treatment. Buf-Vereijken and Wetzels [22] showed that 11 out of 15 patients who required a second course of cyclophosphamide either due to a relapse of their MN or a decline in renal function, had improved outcomes with proteinuria and renal function with few adverse events that were all manageable with dose reductions. Calcineurin inhibitors would not be an attractive option for patients with advanced kidney disease, and therefore, choice is limited with combined steroids and alkylating agent therapy or B-cell depleting agents for this subset of patients.

Data from the Toronto registry show that 9% of patients achieving a state of PR will develop ESRD within 67 months compared to 29% of patients who were nonresponders will develop ESRD within 36 months [23]. In comparison, in our study, patients had an improvement in their eGFR by 12 mL/min during follow-up. We observed >40% of these patients reached a state of CR. There was an initial drop in eGFR by an average of 6 mL/min during the first 3 months of treatment, following which renal function gradually improved. For the 2 patients who did not respond to treatment, one started dialysis 9 years after immunosuppressive therapy. Time course and trend in renal dysfunction in this patient suggest that the need for dialysis may have been delayed although therapy did not improve the degree of proteinuria.

Previous evidence has shown that the level of anti-PLA₂R-Ab corresponds with the time to remission of proteinuria, and patients with higher antibody titer will have a delayed clinical response compared to patients with lower antibody levels [24]. That has been examined for patients with an average eGFR>70 mL/min/1.73 m². Data from our cohort with low eGFR are in line with these findings but there is a trend toward an earlier immunological response. 9 out of 12 patients with antibody testing had high antibody titers; time to achievement of PR was 6 versus 14 months in the previously cited paper [24].

Adverse events needing hospitalization and subsequent development of cancers remain a challenge in these patients. However, age may well be an important determinant of these events. In this cohort, median age was 68 years compared to 59 years in Zonozi et al.’s [5], 60.7 years in Hanset et al.’s [6], 38 years in Jha et al.’s [15], and 48 years in Ponticelli et al.’s [14] study. This study argues that such patients may well benefit from immunological and clinical remission and should not necessarily be excluded from treatment based on eGFR alone. In this study, we note that patients initiated immunosuppressive therapy within a median of 3.5 (IQR 2.3–6.6) months after biopsy diagnosis, during which time there was a decline in eGFR of 11 mL/min. Thirteen patients in Hanset et al.’s [6] study 13 presented with an average eGFR of 48 mL/min and had a progressive drop in renal function to 18 mL/min at the time of receiving rituximab (median of 17 months from presentation). We speculate that monitoring while “waiting for spontaneous remission” may be detrimental for some patients and better prediction algorithms are needed to help identify such patients earlier in the course of their disease. Fourteen patients in this cohort received ImS within 12 months of diagnosis. These studies cannot make the findings applicable for patients with declining renal function over several years.

We acknowledge several limitations of our study including potential selection bias, relatively small number of patients, and lack of a control group. However, this study presents clinical outcome data over a long period of follow-up in a specialist center. All patients had incident disease and we did not include relapses in this study. Excluding patients on non-cyclophosphamide-based ImS therapy has limited the number of patients yet has limited results heterogeneity given the lack of control group. This study challenges the notion of supportive therapy alone in this subset of very high-risk patients in advanced stages of the disease. The study results indicate a significant impact on CKD care and kidney health outcome for this patient subset with PMN. Lastly, majority of patients were Caucasians and men. Therefore, further data would be required to extrapolate these findings to other patient groups. These studies make a strong case for further prospective research into the use of these, and other newer agents to help improve outcomes in delaying progression to ESKD.

In conclusion, this study demonstrates that in patients with PMN and severe renal dysfunction a combination therapy with cyclophosphamide and steroids is safe, and highly effective in achieving both immunological and clinical remission.

An abstract of the present work has been accepted as a mini-oral abstract in the 58th ERA-EDTA Congress, June 5–8, 2021, Berlin, and virtual.

This study protocol was reviewed and approved by Manchester Renal Biobank (IRAS application number 193564, MFT PIN R01011, amendment 2). Manchester Biobank study consents were signed by all patients. Patients consented for publication of the details of their medical cases and any accompanying images.

The authors have no conflict of interest to declare.

No funding source relevant to our study to declare.

Omar Ragy, Patrick Hamilton, and Durga Kanigicherla contributed to study design. Patrick Hamilton was responsible for statistical analysis. Omar Ragy, Durga Kanigicherla, Adil Ahmed, and Anjali Pathi contributed to data base management. Omar Ragy drafted the manuscript; Sandip Mitra and Durga Kanigicherla critically revised the final draft. Omar Ragy, Patrick Hamilton, Sandip Mitra, and Durga Kanigicherla contributed to the interpretation of the results and approved the final version of the paper.

All data generated or analyzed during this study are included in this article and its online supplementary material. Further inquiries can be directed to the corresponding author.