Low-Dose Rituximab vs. Modified Ponticelli Regimen in the Treatment of Primary Membranous Nephropathy – A Randomized Controlled Trial

Introduction

The choice of immunosuppressive therapy for primary membranous nephropathy (PMN) remains a matter of debate.¹ The 2012 KDIGO guidelines recommended the modified Ponticelli regimen (mPR) consisting of alternating months of steroids and cyclophosphamide as first-line therapy.² However, the side effects of both drugs make it less attractive.³ Calcineurin inhibitors (CNIs), although effective in inducing remission, are associated with high relapse and nephrotoxicity rates.^4,5 Rituximab (RTX) has been shown to induce remission in 2/3 patients with MN.^6,7 While some studies have shown a better safety profile for RTX, there is a paucity of prospective head-to-head trials comparing the efficacy of the two regimens.^8,9 There is also uncertainty over the optimum dosing protocol for RTX.² Dosing strategies attempted in previous studies included 375 mg/m² of RTX weekly for four doses or 1 g on day 1 and 15.^10,11 By measuring the circulating CD19⁺ or CD20⁺ B-cells and targeting B-cell depletion to <5 cells/mm³, lower dose strategies have been developed, while reducing cost and maintaining efficacy.^12,13 A recent observational study in India showed 66.7% remission with a total 1g RTX dose.¹⁴ This lower dosing protocol has not been tested in a randomized trial against the mP. Our study aimed to compare a dosing schedule of RTX (500 mg) given on days 1 and 15 with CD19⁺ B-cell monitoring done at the end of the 1^st and 6^th months with cyclical steroids and cyclophosphamide. It was done as a pilot trial to study the feasibility of a larger trial to determine the safety and efficacy of this RTX dose.

Materials and Methods

This study was designed as an open-label RCT comparing the efficacy of RTX with the mPr for PMN treatment [Figure 1]. Approval for the conduct of the study was obtained from the Institutional Review Board (IRB) (EC/525/2018). The trial was registered in ISRCTN (International Standard Randomized Controlled Trial Number) registry (ISRCTN17242711). The study was reported in accordance with the CONSORT 2010 guidelines. The study was conducted in a tertiary care hospital in Western India. Enrolment period for the study was from August 2018 to January 2021 with follow-up till January 2022 for a total study of 42 months.

Close

Figure 1:

Study protocol. PLA2R: Phospholipase A2 receptor, ACEI: Angiotensin converting enzyme inhibitor, ARB: Angiotensin receptor blocker.

Export to PPT

All patients ≥ 18 years of age with biopsy-proven primary MN within 2 years of enrolment, having nephrotic range proteinuria, that is 24-hour urine protein ≥3.5 g or UPCR (urine protein:creatinine ratio) ≥3500 mg/g were included in the study after obtaining informed consent. These patients were positive for anti-PLA2R antibodies by EUROIMMUN Anti-PLA2R (IgG) ELISA (≥14 RU/mL) or their biopsy showed positive PLA2R staining on immunohistochemistry. Patients who tested negative for these were included after ruling out all secondary causes. Angiotensin Converting Enzyme Inhibitor (ACEI) or Angiotensin Receptor Blocker (ARB) were administered for at least 3 months before enrolment if tolerated. Patients showing severe disabling symptoms related to nephrotic syndrome or severe hypoalbuminemia (<2 g/dL) were initiated on immunosuppression before completion of this 3-month period, at the discretion of the treating nephrologist.

Patients with secondary MN, active serious infections, pregnant women, suspected or known hypersensitivity to either interventional drug, and those with persistently low estimated GFR <30 mL/min/1.73m² (as measured by the CKD-EPI equation) in the absence of acute causes were excluded from the study.^15,16 Patients who had received any immunosuppressive therapy in the 4 weeks prior to randomization were excluded.

After obtaining informed consent, patients were randomized with an equal allocation ratio (1:1) to intervention with RTX or the mPr. We used a random number-producing algorithm for block randomization using sealed envelope online software with generation of random allocation sequence, patient enrolment, and assignment to interventions done by the primary investigator.¹⁷

Study arm: RTX injection (500 mg) IV given on days 1 and 15. Each patient received premedication with injection methylprednisolone (125 mg), injection Pheniramine (45.5 mg), and tablet Paracetamol (500 mg).

Comparison arm received Cyclical therapy with 1g IV methylprednisolone daily (Days 1-3), then oral prednisolone (0.5 mg/kg/day) for 27 days (Days 4-30) on months 1, 3, and 5, alternating with oral cyclophosphamide (2.0 mg/kg/day) for 30 days on months 2, 4, and 6.

In the rituximab group, CD19 positive B-cell count was done by flow cytometry at the end of the 1^st and 6^th months. All patients in both arms received trimethoprim-sulfamethoxazole prophylaxis for 6 months. In case of limited or non-response at 6 months, THE decision to change immunosuppression was made by the primary treating nephrologist depending on clinical status and anti-PLA2R titers of the patient.

Primary end point was the proportion of patients reaching complete or partial remission at 6 months, defined according to the 2012 KDIGO guidelines:¹⁸ Complete remission was defined as reduction of proteinuria to <0.3 g/24h (UPCR < 300 mg/g) with normal serum albumin and creatinine. Partial remission was defined as a reduction of proteinuria to <3.5 g/day (UPCR < 3500 mg/g) and ≥50% reduction from peak values with improvement or normalization of serum albumin and stable creatinine.

The secondary end points were: a) the proportion of patients with limited response (proteinuria reduced from baseline by >50% but still >3.5 g/24 h) and no response (reduction of proteinuria <50% from baseline) at 6 and 12 months, b) the rate of relapse, which was defined as reappearance of proteinuria >3.5 g/24 h and at least 50% higher than the lowest post-treatment value in those who previously had remission, c) a trend of change in laboratory parameters, and d) the rate of adverse events during the study period.

Results

Among 91 potentially eligible patients with biopsy-proven membranous nephropathy evaluated, 68 patients were randomized on a 1:1 basis with 34 patients in each study arm [Figure 2].

Close

Figure 2:

Study flow diagram.

Export to PPT

At least 2 months of therapy were completed in all patients in the mPR arm, and all except one received 2 RTX doses (500 mg) in the RTX arm. One patient received only RTX (500 mg) as he had severe low backache 1 week after infusion. Three patients discontinued mPR before 6 months due to adverse effects such as leucopenia and infections. Two patients and one patient, respectively, in each arm were lost to follow-up prior to completion of 6 months, and their remission status could not be ascertained. The median follow-up in the RTX arm was 18 (Interquartile range, IQR = 12-24) months and in the mPR arm, it was 15 (IQR = 6-25) months (p = 0.435).

The baseline characteristics in the two groups have been shown in Table 1. The median UPCR was 8.77 (IQR = 5.95-11.47) g/g in the RTX group and 8.22 (IQR = 5.61-11.37) g/g in the mPR group (p = 0.695). UPCR more than 8 g/g was noted in 20 (58.8%) patients in the rituximab group compared to 18 (52.9%) patients in the mPR group. Severe hypoalbuminemia < 2.5 g/dL was noted in 21 (61.8%) and 18 (58.8%) patients, and anti-PLA2R titers of >250 IU/L were found in 11 (32.4%) and nine (26.5%) patients in the two groups, respectively. Chronic changes, such as glomerulosclerosis, interstitial fibrosis, and tubular atrophy, were seen in 39.4% of biopsies in the RTX arm compared to 55.8% in the mPR arm (p = 0.145). In the RTX arm, 14 patients (41.2%) had received some form of immunosuppression prior to presentation at our center, in comparison to nine patients (26.5%) in the mPR group. Among the RTX group, 7/14 patients had received an incomplete course of mPR, three received tacrolimus, two cyclosporine, and two MMF. Among the mPR group patients, one had received a partial course of mPR, one was given tacrolimus, and four were given MMF. All 23 of these patients had taken steroids at variable doses.

The primary outcome of remission, either complete or partial, at 6 months, as analyzed on an intention-to-treat basis, was achieved in 15/34 patients in the RTX arm (44.12%) compared to 17/34 patients in the mPR arm (50%) giving a risk difference of 5.88% in favor of mPR (95% CI -29.57% to 17.80%, p = 0.626) [Figure 3]. Patients who transitioned to alternative therapies before 6 months were categorized as resistant to the primary therapy and included in the intention-to-treat analysis as non-responders.

Close

Figure 3:

Complete or partial remission at 6 months for rituximab vs mPR demonstrated as % risk difference. Primary outcome of complete or partial remission at 6 months was analysed on an intention-to-treat basis while the remission status at 12 months was analysed on per-protocol basis. CR/PR: Complete remission or partial remission.

Export to PPT

The secondary outcome was the rates of complete and partial remission, as assessed by a per-protocol analysis [Table 2]. Complete data at 6 and 12 months, after exclusion of deaths and loss to follow-up, were available for 30 and 27 patients, respectively, in the RTX arm and for 29 and 23 patients, respectively, in the mPR arm.

At 12 months, 16/27 (59.3%) patients in the RTX arm and 16/23 (69.6%) patients in the mPR arm had either complete or partial remission (RD -10.3%, 95% CI -36.7% to 16.1%, p = 0.444). The mean time to 50% reduction in proteinuria was 3.6 (±3.7) months in the RTX arm and 3.3 (±2.6) months in the mPR arm (p = 0.778). The time to partial and complete remission was 5.9 (±3.7) and 12.6 (±7.0) months, respectively, in the RTX group compared to 4.7 (±3.1) [p = 0.264] and 8.1 (±7.3) months [p = 0.247] in the mPR group.

The median UPCR at 6 months was lower in the mPR group compared to the RTX group (2.24 g/g vs. 4.19 g/g, p = 0.097) [Table 3]. The mean decline in anti-PLA2R titers at 6 months was higher in the mPR group (338.89 RU/mL) compared to the RTX group (79.65 RU/mL) (p = 0.080).

Of 20 patients who achieved remission anytime in the RTX arm, five had a relapse (20%) at a mean duration of 12.4±10.2 months after initial remission. Of 21 patients in the mPR arm who had achieved remission, four (19.05%) relapsed at a mean duration of 12±13.9 months (p = 0.645).

Immunosuppression changes occurred in six patients (17.64%) in the RTX arm compared to four (11.76%) in the mPR arm, with average time taken to change therapy being 8.7±2.9 months and 6±0 months, respectively. Five patients in the RTX arm crossed over to mPR due to non-response, two of whom achieved remission. In the mPR arm, two patients each were started on CNIs and mycophenolate.

Subgroup analysis on the efficacy of RTX vs. mPr was done according to baseline characteristics known to influence outcome [Figure 4]. Significant differences in efficacy across age groups, sex, and baseline proteinuria were not found. The subgroup with antiPLA2R titers > 150 RU/mL had higher odds of remission with mPR, but the difference was not statistically significant.

Close

Figure 4:

Subgroup analysis for primary outcome of complete or partial remission at 6 months. PLA2R: Phospholipase A2 receptor, UPCR: Urine protein creatinine ratio.

Export to PPT

All the patients in the RTX arm had significant B-cell depletion to < 1% of the total leucocytes. The median CD19 +ve B-cell count was 0 (Interquartile range [IQR] = 0 to 2.1) cells/mm³. As the percentage of leucocytes, the median value was 0% (IQR = 0 to 0.1%) at 1 month. The median CD19 +ve B-cell count at 6 months post-RTX therapy was 91.2 (IQR = 4.1 to 225.3) cells/mm³, with 64% having B-cell count recovery. There was no significant difference in the odds of remission among those with persistently low CD19 +ve cell count <1% at 6 months (OR 3.0, 95% CI 0.440 to 20.436) and those with cell count recovery (OR 0.600, 95% CI 0.275 to 1.309, p = 0.198).

The number of adverse events was significantly higher in the mPR group at 56 events in 22 (64.71%) patients, respectively, compared to 22 events (p = 0.007) in 13 (38.24%) patients (p = 0.024) in the RTX group [Table 4]. In the mPR arm, 11 (32.35%) patients had immunosuppression interruption, most of which were due to infections or leucopenia.

There were significantly more infections in the mPR arm at a rate of 86 per 100 patients compared to 46 per 100 patients in the RTX group (p = 0.039). The time to onset of first infection after protocol therapy averaged 4.5 months in the RTX group in contrast to 2.9 (±2.36) months in the mPR group.

In the mPR group, nine patients (30.8%) developed leucopenia, of whom four had an infective episode. Neutropenia was noted in three patients, all of whom required G-CSF (Granulocyte-Colony Stimulating Factor) injections, and two of these patients had life-threatening febrile neutropenia, among whom one died due to sepsis. The average time of onset of leucopenia was 3.2 (±1.3) months after mPr initiation.

Infusion reactions occurred only in the RTX arm, affecting 2/34 patients. Both were mild episodes with itching and urticarial reaction, which subsided immediately with antihistaminic and hydrocortisone injection.

Patient-reported cosmetic side effects of steroids, such as Cushingoid features and acne, occurred in six patients (17.65%) of the mPR arm. The onset of impaired glucose tolerance or diabetes mellitus was significantly more common in the mPR group, occurring in seven patients (20.59%), compared to one patient in the RTX arm (p = 0.019).

AKI or a progressive rise in creatinine occurred in nine patients (26.5%) in the RTX arm and seven patients (20.6%) in the mPR arm (p = 0.567). In each arm, one patient developed ESKD, both at 24 months after the start of therapy.

There were three deaths in the RTX group and four in the mPR group during follow-up. The average time to death after the protocol therapy was 8 ±5.1 months in the rituximab group compared to 3.8±1.5 months in the mPR group (p = 0.086). There was one death in each arm due to tuberculosis. In the mPR group, one succumbed to pneumonia and sepsis. The others expired in their hometown due to unknown causes. None of these seven patients had achieved remission.

Discussion

This pilot study demonstrates the feasibility of conducting a randomized trial comparing RTX with the mPr in India. At 6 months, nearly 50% of patients achieved clinical remission, with RTX showing a favorable safety profile. While the short-term remission rates were comparable, a multicentric study would be required to assess the non-inferiority of these treatment modalities. In terms of disease severity markers, the two groups performed similarly, with a modest advantage in the modified Ponticelli group. The decline in creatinine, proteinuria, and antiPLA2R titers and the rise in eGFR were greater in the mPR group compared to RTX. There was no significant difference in outcomes in subgroup analysis comparing RTX with mPR, given the small sample size. A larger study focusing on those with anti-PLA2R titers >150 RU/mL would be required to assess the efficacy of RTX in this subgroup.

RTX is expensive, and several studies have shown that lower doses may be able to achieve adequate CD19 +ve B-cell suppression.^12,13,19 We used half the dose used in the MENTOR trial (initial 2 g dose), and less than that of the GEMRITUX trial (∼1.45 g) was used.^20,21 All patients exhibited depletion of B cells at 1 month. The study provided evidence of efficacy, with appropriate short-term suppression of B-cell numbers achieved using lower RTX doses. Our findings at 6 months align with those reported in other pivotal studies, such as GEMRITUX, MENTOR, and RI-CYCLO, which evaluated remission outcomes in membranous nephropathy. The GEMRITUX and MENTOR trials demonstrated a 35% remission rate at 6 months with RTX, which is comparable to the remission rate observed in our study.^20,21 In the RI-CYCLO study, the remission rates at 6 months were 53% and 64% in the RTX and mPR arms, respectively, which were slightly higher than the rates observed in our study.²² In the MENTOR and RI-CYCLO trials, remission outcomes were more evident by 12 months with 60% and 66% of patients in the rituximab arm having achieved remission. In our study, by 12 months of follow-up, 59% of patients in the RTX arm achieved remission; however, beyond 6 months, the results were observational, as treatment was guided by the primary nephrologist rather than a protocolized approach.

RTX was associated with a lower risk of drug-related adverse events, including serious adverse events, compared to the mPr. More patients in the mPR arm required hospital admission and interruption of immunosuppression, particularly due to infections and cytopenias, which had significant clinical implications. Steroid-related adverse events, such as cosmetic side effects and hyperglycemia, were also more frequent in the mPR group. All patients receiving RTX were pre-medicated. The two observed infusion reactions were mild, non-life-threatening, and did not necessitate discontinuation of therapy. Notably, adverse event rates with RTX were 71% and 43% in the MENTOR and RI-CYCLO trials, respectively.^21,22 While the RI-CYCLO trial reported comparable adverse event rates between RTX and mPR arms, the lower RTX dose was associated with fewer adverse events compared to mPR.²²

All seven deaths were of non-responders to therapy. The 2/3 deaths in the RTX group and all four in the mPR group occurred during the 6-month therapeutic period. The 10% mortality rate in our study cohort was higher than that seen in most other PMN studies.^21,23 It may be related to the greater disease severity in our PMN cohort. This study was conducted at a nephrology referral center in Western India. The disease severity at presentation was worse than in other major therapeutic PMN trials.^23,24 Severe hypoalbuminemia <2.5 g/dL, proteinuria >8 g, and mean anti-PLA2R titers >250 RU/mL were noted in both groups, with a significant proportion having chronic changes on biopsy. Nearly 1/3 of the cohort had AKI prior to therapy initiation, and 1/4 had pre-existing comorbidities such as hypertension and diabetes. Multiple factors beyond immunosuppressive therapy could contribute to kidney function decline, including baseline disease severity, chronic structural changes identified on kidney biopsy, pre-existing comorbidities like hypertension and diabetes, and AKI episodes before treatment initiation.

This study had certain limitations. It was not blinded. As a pilot study, it was not able to reach the estimated sample size for evaluating the study hypothesis. The follow-up of a few patients was severely hampered by the COVID-19 pandemic in 2020. One limitation was the higher proportion of serum anti-PLA2R positivity in the RTX arm, which could have influenced outcomes. While PLA2R positivity is linked to active disease, the lower incidence of positive cases in the mPR arm could indicate the onset of spontaneous immunological remission. Further, discrepant results in biopsy PLA2R positivity were also noted. Future studies may need to stratified randomization based on PLA2R positivity to avoid this potential bias. Another drawback was that a few of the patients had received some therapy prior to presentation. However, patients in this real-world study represent those presenting to a nephrology referral center with severe PMN and having received some form of immunosuppression. After 6 months, therapy was continued at the discretion of the treating nephrologist, and further follow-up was observational. Protocolized administration of further doses of RTX or crossover to cyclical steroids and cyclophosphamide at 6 months, depending on response, would be required to assess long-term outcomes.

While this pilot study demonstrates promising preliminary outcomes, we recognize that a longer follow-up period is essential to draw more robust conclusions regarding the sustained effectiveness and potential long-term safety of this protocol. Previous studies, such as the GEMRITUX and MENTOR trials, with an extended follow-up, assessed the durability of remission and the stability of kidney function. The 6-month follow-up in our study limits our ability to assess long-term kidney outcomes and relapse potential. However, our results suggest an initial trend towards comparable remission rates that warrants further investigation. Although observational data from patients who have completed 12 months of follow-up indicate similar trends, further research with a larger sample size and extended follow-up is necessary to confirm these findings. In conclusion, our study provides valuable early insights into the comparative efficacy of low-dose RTX and mPR in PMN and lays the groundwork for future research with longer follow-up using different dosing protocols of RTX to determine sustained remission, relapse rates, and kidney function stability.