The STARMEN trial indicates that alternating treatment with corticosteroids and cyclophosphamide is superior to sequential treatment with tacrolimus and rituximab in primary membranous nephropathy

A cyclical corticosteroid-cyclophosphamide regimen is recommended for patients with primary membranous nephropathy at high risk of progression. We hypothesized that sequential therapy with tacrolimus and rituximab is superior to cyclical alternating treatment with corticosteroids and cyclophosphamide in inducing persistent remission in these patients. This was tested in a randomized, open-label controlled trial of 86 patients with primary membranous nephropathy and persistent nephrotic syndrome after six-months observation and assigned 43 each to receive six-month cyclical treatment with corticosteroid and cyclophosphamide or sequential treatment with tacrolimus (full-dose for six months and tapering for another three months) and rituximab (one gram at month six). The primary outcome was complete or partial remission of nephrotic syndrome at 24 months. This composite outcome occurred in 36 patients (83.7%) in the corticosteroid-cyclophosphamide group and in 25 patients (58.1%) in the tacrolimus-rituximab group (relative risk 1.44; 95% confidence interval 1.08 to 1.92). Complete remission at 24 months occurred in 26 patients (60%) in the corticosteroid-cyclophosphamide group and in 11 patients (26%) in the tacrolimus-rituximab group (2.36; 1.34 to 4.16). Anti-PLA2R titers showed a significant decrease in both groups but the proportion of anti-PLA2R-positive patients who achieved immunological response (depletion of anti-PLA2R antibodies) was significantly higher at three and six months in the corticosteroid-cyclophosphamide group (77% and 92%, respectively), as compared to the tacrolimus-rituximab group (45% and 70%, respectively). Relapses occurred in one patient in the corticosteroid-cyclophosphamide group, and three patients in the tacrolimus-rituximab group. Serious adverse events were similar in both groups. Thus, treatment with corticosteroid-cyclophosphamide induced remission in a significantly greater number of patients with primary membranous nephropathy than tacrolimus-rituximab. A cyclical corticosteroid-cyclophosphamide regimen is recommended for patients with primary membranous nephropathy at high risk of progression. We hypothesized that sequential therapy with tacrolimus and rituximab is superior to cyclical alternating treatment with corticosteroids and cyclophosphamide in inducing persistent remission in these patients. This was tested in a randomized, open-label controlled trial of 86 patients with primary membranous nephropathy and persistent nephrotic syndrome after six-months observation and assigned 43 each to receive six-month cyclical treatment with corticosteroid and cyclophosphamide or sequential treatment with tacrolimus (full-dose for six months and tapering for another three months) and rituximab (one gram at month six). The primary outcome was complete or partial remission of nephrotic syndrome at 24 months. This composite outcome occurred in 36 patients (83.7%) in the corticosteroid-cyclophosphamide group and in 25 patients (58.1%) in the tacrolimus-rituximab group (relative risk 1.44; 95% confidence interval 1.08 to 1.92). Complete remission at 24 months occurred in 26 patients (60%) in the corticosteroid-cyclophosphamide group and in 11 patients (26%) in the tacrolimus-rituximab group (2.36; 1.34 to 4.16). Anti-PLA2R titers showed a significant decrease in both groups but the proportion of anti-PLA2R-positive patients who achieved immunological response (depletion of anti-PLA2R antibodies) was significantly higher at three and six months in the corticosteroid-cyclophosphamide group (77% and 92%, respectively), as compared to the tacrolimus-rituximab group (45% and 70%, respectively). Relapses occurred in one patient in the corticosteroid-cyclophosphamide group, and three patients in the tacrolimus-rituximab group. Serious adverse events were similar in both groups. Thus, treatment with corticosteroid-cyclophosphamide induced remission in a significantly greater number of patients with primary membranous nephropathy than tacrolimus-rituximab. Primary membranous nephropathy (PMN) is one of the most common causes of nephrotic syndrome in adults.1Couser W. Primary membranous nephropathy.Clin J Am Soc Nephrol. 2017; 12: 983-997Crossref PubMed Scopus (245) Google Scholar In 70%–80% of cases, the disease is mediated by autoantibodies targeting the phospholipase A2 receptor (PLA2R) expressed in podocytes and in 3%–5% by autoantibodies to thrombospondin type 1 domain–containing 7A (THSD7A).2Beck Jr., L.H. Bonegio R.G. Lambeau G. et al.M-Type phospholipase A2 receptor as target antigen in idiopathic membranous nephropathy.N Engl J Med. 2009; 361: 11-21Crossref PubMed Scopus (1419) Google Scholar, 3Ronco P. Debiec H. Molecular pathogenesis of membranous nephropathy.Annu Rev Pathol. 2020; 15: 287-313Crossref PubMed Scopus (42) Google Scholar, 4Tomas N.M. Beck Jr., L.H. Meyer-Schwesinger C. et al.Thrombospondin type-1 domain–containing 7A in idiopathic membranous nephropathy.N Engl J Med. 2014; 371: 2277-2287Crossref PubMed Scopus (496) Google Scholar Spontaneous remission occurs in one-third of patients,5Polanco N. Gutiérrez E. Covarsí A. et al.Spontaneous remission of nephrotic syndrome in idiopathic membranous nephropathy.J Am Soc Nephrol. 2010; 21: 697-704Crossref PubMed Scopus (248) Google Scholar and therefore an observational period of at least 6 months is recommended.6Waldman M. Austin H.A. Treatment of idiopathic membranous nephropathy.J Am Soc Nephrol. 2012; 23: 1617-1630Crossref PubMed Scopus (62) Google Scholar, 7Hofstra J.M. Fervenza F.C. Wetzels J.F.M. Treatment of idiopathic membranous nephropathy.Nat Rev Nephrol. 2013; 9: 443-458Crossref PubMed Scopus (88) Google Scholar, 8Kidney Disease: Improving Global Outcomes (KDIGO) Glomerulonephritis Work GroupKDIGO clinical practice guidelines for glomerulonephritis.Kidney Int Suppl. 2012; 2 (139–27)Google Scholar Conversely, about 50% of cases with persistent nephrotic syndrome eventually progress to end-stage kidney disease, and immunosuppressive therapy is recommended for these patients. Controversy persists about the most effective type of immunosuppressive regimen. The 2012 Kidney Disease: Improving Global Outcomes (KDIGO) guidelines for glomerulonephritis recommended a 6-month cyclic regimen of alternating alkylating agents (usually cyclophosphamide) plus corticosteroids for patients at high risk of progression, since it was the only regimen that was shown to be effective in preventing end-stage kidney disease.8Kidney Disease: Improving Global Outcomes (KDIGO) Glomerulonephritis Work GroupKDIGO clinical practice guidelines for glomerulonephritis.Kidney Int Suppl. 2012; 2 (139–27)Google Scholar, 9Ponticelli C. Zucchelli P. Passerini P. et al.A randomized trial of methylprednisolone and chlorambucil in idiopathic membranous nephropathy.N Engl J Med. 1989; 320: 8-13Crossref PubMed Scopus (240) Google Scholar, 10Ponticelli C. Zucchelli P. Passerini P. et al.A 10-year follow-up of a randomized study with methylprednisolone and chlorambucil in membranous nephropathy.Kidney Int. 1995; 48: 1600-1604Abstract Full Text PDF PubMed Scopus (336) Google Scholar, 11Ponticelli C. Altieri P. Scolari F. et al.A randomized study comparing methylprednisolone plus chlorambucil versus methylprednisolone plus cyclophosphamide in idiopathic membranous nephropathy.J Am Soc Nephrol. 1998; 9: 444-450Crossref PubMed Google Scholar, 12Jha V. Ganguli A. Saha T.K. et al.A randomized, controlled trial of steroids and cyclophosphamide in adults with nephrotic syndrome caused by idiopathic membranous nephropathy.J Am Soc Nephrol. 2007; 18: 1899-1904Crossref PubMed Scopus (215) Google Scholar However, given the important number of serious adverse events associated with cumulative doses of alkylating agents, treatment alternatives were introduced. Calcineurin inhibitors (both cyclosporine and tacrolimus) have shown efficacy in inducing remission of nephrotic syndrome in about 70% of patients.13Cattran D.C. Appel G.B. Hebert L.A. et al.Cyclosporine in patients with steroid-resistant membranous nephropathy: a randomized trial.Kidney Int. 2001; 59: 1484-1490Abstract Full Text Full Text PDF PubMed Scopus (297) Google Scholar,14Praga M. Barrio V. Juárez G.F. Luño J. Tacrolimus monotherapy in membranous nephropathy: a randomized controlled trial.Kidney Int. 2007; 71: 924-930Abstract Full Text Full Text PDF PubMed Scopus (216) Google Scholar However, the main limitation of these drugs is the high rate of relapse after discontinuation. An observational study found a reduction in relapse rates when rituximab was administered at the time of tapering off cyclosporine or tacrolimus,15Segarra A. Praga M. Ramos N. et al.Successful treatment of membranous glomerulonephritis with rituximab in calcineurin inhibitor-dependent patients.Clin J Am Soc Nephrol. 2009; 4: 1083-1088Crossref PubMed Scopus (82) Google Scholar and a pilot study reported encouraging results of a combined therapy with cyclosporine plus rituximab in high-risk PMN patients.16Waldman M. Beck Jr., L.H. Braun M. et al.Membranous nephropathy: pilot study of a novel regimen combining cyclosporine and Rituximab.Kidney Int Rep. 2016; 1: 73-84Abstract Full Text Full Text PDF PubMed Scopus (23) Google Scholar More recently, the efficacy of rituximab monotherapy has received great attention.17Ruggenenti P. Cravedi P. Chianca A. et al.Rituximab in idiopathic membranous nephropathy.J Am Soc Nephrol. 2012; 23: 1416-1425Crossref PubMed Scopus (188) Google Scholar A single dose of rituximab was suggested to be effective for induction of remission in an observational cohort,18Cravedi P. Ruggenenti P. Sghirlanzoni M.C. Remuzzi G. Titrating rituximab to circulating B cells to optimize lymphocytolytic therapy in idiopathic membranous nephropathy.Clin J Am Soc Nephrol. 2007; 2: 932-937Crossref PubMed Scopus (124) Google Scholar although in recent clinical trials, higher doses were needed for optimal efficacy.19Dahan K. Debiec H. Plaisier E. et al.Rituximab for severe membranous nephropathy: a 6-month trial with extended follow-up.J Am Soc Nephrol. 2017; 28: 348-358Crossref PubMed Scopus (177) Google Scholar, 20Fervenza F.C. Appel G.B. Barbour S.J. et al.Rituximab or cyclosporine in the treatment of membranous nephropathy.N Engl J Med. 2019; 381: 36-46Crossref PubMed Scopus (157) Google Scholar, 21Seitz-Polski B. Dahan K. Debiec H. et al.High-dose rituximab and early remission in pla2r1-related membranous nephropathy.Clin J Am Soc Nephrol. 2019; 14: 1173-1182Crossref PubMed Scopus (43) Google Scholar Indeed, the superior efficacy of rituximab versus cyclosporine in the recent MENTOR (Membranous Nephropathy Trial of Rituximab) study was achieved using a total dose of 4 g rituximab.20Fervenza F.C. Appel G.B. Barbour S.J. et al.Rituximab or cyclosporine in the treatment of membranous nephropathy.N Engl J Med. 2019; 381: 36-46Crossref PubMed Scopus (157) Google Scholar The need for head-to-head trials comparing the 6-month cyclic alternating treatment with corticosteroids and cyclophosphamide with the newer therapeutic alternatives (calcineurin inhibitors, rituximab) was stressed in a recent KDIGO conference.22Floege J. Barbour S.J. Cattran D.C. et al.Management and treatment of glomerular diseases (part 1): conclusions from a Kidney Disease: Improving Global Outcomes (KDIGO) Controversies Conference.Kidney Int. 2019; 95: 268-280Abstract Full Text Full Text PDF PubMed Scopus (139) Google Scholar We designed the STARMEN (Sequential Treatment with Tacrolimus and Rituximab Versus Alternating Corticosteroids and Cyclophosphamide in PMN) study to compare cyclic alternating treatment of corticosteroids and cyclophosphamide with a sequential treatment of tacrolimus and rituximab in the induction and maintenance of nephrotic syndrome remission for up 24 months. In addition, we studied the occurrence of relapses after remission and the role of anti-PLA2R autoantibodies over the course of the treatment. From June 2014 through June 2017, 130 patients were assessed for eligibility, of whom 44 (33%) were excluded from the study. Main reasons for screening failure were not meeting the eligibility criteria (21 patients), unwillingness to participate in the study (14 patients), and diagnosis of a secondary cause of membranous nephropathy (9 patients). The remaining 86 patients who met the eligibility criteria were randomly assigned to the corticosteroid-cyclophosphamide group (43 patients) or to the tacrolimus-rituximab group (43 patients) (Figure 1). As presented in Table 1, no significant differences between groups were observed at baseline. Baseline stored serum samples were available in 69 patients, and anti-PLA2R was positive in 53 (77%). Sensitivity analyses showed no differences at baseline between anti-PLA2R–positive and anti-PLA2R–negative patients nor between patients with or without anti-PLA2R determinations at baseline (Supplementary Tables S1 and S2). One patient was anti-THSD7A positive. Kidney biopsies were performed 8 (range, 6–18) months before randomization. Seventy-three patients (85%) had de novo PMN and 13 (15%) relapsing PMN (Table 1).Table 1Baseline characteristics according to treatment groupCharacteristicAll patients (n = 86)Corticosteroid-cyclophosphamide (n = 43)Tacrolimus-rituximab (n = 43)P valueAge, yr55.7 ± 11.456.2 ± 12.055.2 ± 10.81Male sex55 (64)24 (55)31 (72)0.12De novo/relapsing PMN73 (85) / 13 (15)36 (83)/7 (17)37 (86)/6 (14)0.70Weight, kg78.5 ± 16.076.9 ± 17.080.0 ± 15.00.37Blood pressure, mm Hg Systolic128.3 ± 15.9129.4 ± 17.7127.1 ± 14.00.25 Diastolic76.0 ± 9.875.1 ± 10.777.0 ± 8.80.37Serum creatinine, mg/dl1.0 ± 0.31.0 ± 0.31.0 ± 0.280.48eGFR, ml/min per 1.73 m2aeGFR was calculated according to the Chronic Kidney Disease Epidemiology Collaboration equation.79.8 ± 23.579.1 ± 25.580.5 ± 21.60.78Serum albumin, g/dl2.6 (2.2–2.9)2.6 (2.3–2.9)2.6 (2.0–2.9)0.47Serum cholesterol, mg/dl263.9 ± 64.0264.1 ± 57.8263.8 ± 70.20.34Anti-PLA2R–positive patientsbAnti-PLA2R positivity defined as >14 RU/ml.53 of 69 (77)29 of 37 (78)cIn 6 of 43 patients (16%) anti-PLA2R was not determined at baseline.24 of 32 (75)dIn 11 of 43 patients (25.5%) anti-PLA2R was not determined at baseline.0.38Anti-PLA2R, RU/ml80 (44–149)59 (37–150)113 (61–151)0.1Urinary protein, g/24 h7.4 (5.2–11.5)7.4 (4.8–11.3)7.4 (6.7–11.6)0.22Concomitant treatment ACEis and/or ARBs82 of 83 (99)39 of 43 (91)43 of 43 (100)0.48 Diuretics70 of 83 (84)35 of 43 (81)35 of 43 (81)0.45ACEis, angiotensin-converting enzyme inhibitors; Anti-PLA2R, anti–phospholipase A2 receptor autoantibodies; ARBs, angiotensin II receptor blockers; eGFR, estimated glomerular filtration rate; RU, relative units.Values are presented as mean ± SD, n (%), or median (interquartile range).a eGFR was calculated according to the Chronic Kidney Disease Epidemiology Collaboration equation.b Anti-PLA2R positivity defined as >14 RU/ml.c In 6 of 43 patients (16%) anti-PLA2R was not determined at baseline.d In 11 of 43 patients (25.5%) anti-PLA2R was not determined at baseline. Open table in a new tab ACEis, angiotensin-converting enzyme inhibitors; Anti-PLA2R, anti–phospholipase A2 receptor autoantibodies; ARBs, angiotensin II receptor blockers; eGFR, estimated glomerular filtration rate; RU, relative units. Values are presented as mean ± SD, n (%), or median (interquartile range). All patients received at least 1 month of the assigned therapeutic intervention. Two patients in the corticosteroid-cyclophosphamide group (4.6%) and 6 (13.9%) in the tacrolimus-rituximab group discontinued the intervention (Figure 1). The remaining 41 in the corticosteroid-cyclophosphamide group and 37 in the tacrolimus-rituximab group) received the complete intervention. Patients assigned to the corticosteroid-cyclophosphamide group received a median dose of oral methylprednisolone of 0.49 ± 0.05 mg/kg/day in months 1, 3, and 5, with a total cumulative dose of 3.4 ± 0.9 g. The total cumulative dose of intravenous methylprednisolone was 8.2 ± 1.4 g. The total cumulative dose of cyclophosphamide was 10 ± 3.5 g. Doses and blood levels of tacrolimus in the tacrolimus-rituximab group are presented in Supplementary Table S3. Three patients in this group received a second dose of rituximab (0.5 g in 2 patients, 1 g in 1 patient) at months 12, 12, and 18, respectively, and 2 others received additional doses of tacrolimus beyond month 9. During the follow-up period after the end of the assigned intervention, 5 patients in each group were switched to a nonstudy intervention owing to a lack of efficacy of the assigned treatment. In the corticosteroid-cyclophosphamide group, 2 patients were treated with rituximab at month 14, 2 patients received tacrolimus from month 18, and 1 patient cyclosporine from month 16. In the tacrolimus-rituximab group, the 5 patients received a 6-month cyclic treatment with corticosteroid and cyclophosphamide at month 12 (2 patients), month 14 (2 patients), or month 21 (1 patient). All the patients who were switched to a nonstudy intervention were considered to be nonresponders (Figure 1). Follow-up was complete in all of the 86 enrolled patients. Thirty-six patients (83.7%) in the corticosteroid-cyclophosphamide group and 25 patients (58.1%) in the tacrolimus-rituximab group had a primary outcome of complete/partial remission at 24 months (relative risk [RR] 1.44, 95% confidence interval [CI] 1.08 to 1.92) (Table 2; Figure 2a). The per-protocol analysis confirmed significant differences in the primary outcome between groups: 35 of 41 (85%) in the corticosteroid-cyclophosphamide group and 22 of 37 (59%) in the tacrolimus-rituximab group (RR 1.44, 95% CI 1.07 to 1.93) (Table 2). The difference in the number of patients with complete or partial remission in both groups was already significant at month 3 and was maintained throughout the study (Table 2; Figure 2a).Table 2Composite outcome of complete or partial remission at 3 to 24 monthsIntention-to-treat analysisTime from randomizationCorticosteroid- cyclophosphamide (n = 43)Tacrolimus-rituximab (n = 43)Relative risk (95% CI)3 mo22 of 43 (51)12 of 43 (28)1.83 (1.04–3.22)6 mo32 of 43 (74)19 of 43 (44)1.68 (1.15–2.46)12 mo34 of 43 (79)22 of 43 (51)1.55 (1.11–2.15)18 mo36 of 43 (84)23 of 43 (53)1.57 (1.15–2.13)24 mo36 of 43 (84)25 of 43 (58)1.44 (1.08–1.92)Per-protocol analysisTime from randomizationCorticosteroid- cyclophosphamide (n = 41)Tacrolimus-rituximab (n = 37)Relative risk (95% CI)3 mo22 of 41 (51)12 of 37 (32)1.65 (0.96–2.85)6 mo31 of 41 (75)18 of 37 (48)1.55 (1.07–2.26)12 mo33 of 41 (80)20 of 37 (54)1.49 (1.07–2.08)18 mo35 of 41 (85)20 of 37 (54)1.58 (1.14–2.18)24 mo35 of 41 (85)22 of 37 (59)1.44 (1.07–1.93)CI, confidence interval.The intention-to-treat analyses included all the patients who underwent randomization, whereas the per-protocol analyses included all the patients who received the full course of trial medications. In both analyses, the primary outcome was the composite of complete or partial remission at 24 months.Data are n (%) or 95% CI. Open table in a new tab CI, confidence interval. The intention-to-treat analyses included all the patients who underwent randomization, whereas the per-protocol analyses included all the patients who received the full course of trial medications. In both analyses, the primary outcome was the composite of complete or partial remission at 24 months. Data are n (%) or 95% CI. As shown in Table 3 and Figure 2b, 26 patients (60%) in the corticosteroid-cyclophosphamide group achieved complete remission at 24 months. In the tacrolimus-rituximab group, 11 patients (26%) achieved complete remission at 24 months (RR 2.36, 95% CI 1.34 to 4.16).Table 3Complete remission at 3 to 24 monthsIntention-to-treat analysisTime from randomizationCorticosteroid- cyclophosphamide (n = 43)Tacrolimus-rituximab (n = 43)Relative risk (95% CI)3 mo1 of 43 (2)0 of 43 (0)6 mo6 of 43 (14)0 of 43 (0)12 mo14 of 43 (33)4 of 43 (9)3.50 (1.25–9.78)18 mo19 of 43 (44)7 of 43 (16)2.71 (1.27–5.78)24 mo26 of 43 (60)11 of 43 (26)2.36 (1.34–4.16)Per-protocol analysisTime from randomizationCorticosteroid- cyclophosphamide (n = 41)Tacrolimus-rituximab (n = 37)Relative risk (95% CI)3 mo1 of 41 (2)0 of 37 (0)6 mo5 of 41 (13)0 of 37 (0)12 mo13 of 41 (32)4 of 37 (11)2.93 (1.05–8.21)18 mo18 of 41 (44)7 of 37 (19)2.31 (1.09–4.92)24 mo25 of 41 (61)10 of 37 (27)2.26 (1.26–4.04)CI, confidence interval.The intention-to-treat analyses included all the patients who underwent randomization, whereas the per-protocol analyses included all the patients who received the full course of trial medications. In both analyses, the primary outcome was the composite of complete or partial remission at 24 months.Data are n (%) or 95% CI. Open table in a new tab CI, confidence interval. The intention-to-treat analyses included all the patients who underwent randomization, whereas the per-protocol analyses included all the patients who received the full course of trial medications. In both analyses, the primary outcome was the composite of complete or partial remission at 24 months. Data are n (%) or 95% CI. A tendency for a greater efficacy of the corticosteroid-cyclophosphamide treatment was found across different non-prespecified subgroups defined by baseline values of proteinuria, serum albumin, serum creatinine, anti-PLA2R levels. and age, although no differences were found in female patients (Supplementary Figure S1). When baseline characteristics of patients were compared between those who achieved complete/partial remission and those with no response, a significantly higher proportion of men (80% vs. 57%) and a significantly higher proteinuria were observed among nonresponders (Supplementary Table S4). Proteinuria decreased from a median 7.4 g/24 h (interquartile range 4.8–11.3) at baseline to 0.35 g/24 h (0.2–9) at 24 months in the corticosteroid-cyclophosphamide group and from 7.4 g/24 h (6.7–11.6) at baseline to 1 g/24 h (0.3–3.3) at 24 months in the tacrolimus-rituximab group (between-group difference P = 0.005) (Figure 3a; Supplementary Table S5). Serum albumin increased from a mean 2.6 ± 0.1 g/dl at baseline to 4 ± 0.1 g/dl at 24 months in the corticosteroid-cyclophosphamide group and from 2.6 ± 0.1 g/dl at baseline to 3.9 ± 0.1 g/dl at 24 months in the tacrolimus-rituximab group (between-group difference P = 0.2) (Figure 3b; Supplementary Table S5). There was a nonsignificant trend for higher values of estimated glomerular filtration rate (eGFR) in the corticosteroid-cyclophosphamide group than in the tacrolimus-rituximab group throughout the follow-up (Figure 3c; Supplementary Table S6). At 24 months, the numbers of patients with a ≥50% increases of baseline serum creatinine were 1 (2%) in the corticosteroid-cyclophosphamide group and 5 (12%) in the tacrolimus-rituximab group (P = 0.2). The numbers of patients with preserved renal function (eGFR ≥45 ml/min per 1.73 m2) at 24 months were 40 (93%) in the corticosteroid-cyclophosphamide group and 37 (86%) in the tacrolimus-rituximab group (P = 0.48). The only patient who developed end-stage kidney disease was a 73-year-old man who had been assigned to the corticosteroid-cyclophosphamide group. Three months after treatment completion he was excluded from the study because of persistent massive proteinuria and declining renal function and switched to rituximab (2 doses of 1 g, 15 days apart). No response was observed, and chronic dialysis was initiated 16 months after randomization. Anti-PLA2R levels showed a significant decrease in both groups (Table 4; Figure 3a). The proportions of anti-PLA2R–positive patients who achieved immunologic response at 3 and 6 months were significantly higher in the corticosteroid-cyclophosphamide group (77% and 92%, respectively) than in the tacrolimus-rituximab group (45% and 70%, respectively) (Table 4). Most patients (80%) who achieved immunologic response during the study presented remission of nephrotic syndrome at 24 months. Immunologic response at 3 months (P = 0.036) and 6 months (P = 0.005) was associated with remission at 24 months. Nonresponding patients showed a slower decline in anti-PLA2R levels and a significantly lower proportion of immunologic responses, compared with patients who achieved a complete or partial remission of nephrotic syndrome (Supplementary Table S7).Table 4Evolution of anti-PLA2R antibodies and percentage of immunologic response according to treatment groupTime from randomizationAnti-PLA2R, RU/ml, median (IQR)Immunologic response, %All patientsCorticosteroid-cyclophosphamideTacrolimus-rituximabP valueCorticosteroid-cyclophosphamideTacrolimus-rituximabP valueBaseline80 (44–149)59 (37–150)113 (61–151)0.13 mo2.3 (1.3–65)1.4 (0.7–14)33 (2.6–76)0.00377450.036 mo1.4 (0.7–7.4)1.4 (0–1.7)2.6 (1.4–30)0.0292700.0412 mo1.6 (0–8.1)1.5 (0–5)2.1 (0–15)0.588790.3118 mo1.5 (0–2)1.3 (0–2.3)1.6 (0–15)0.388830.8824 mo1.6 (1.3–5.1)1.4 (0.9–3.2)1.9 (1.7–14)0.0688830.91IQR, interquartile range; PLA2R, phospholipase A2 receptor; RU, relative units.Differences of median were compared with the Mann-Whitney test. Open table in a new tab IQR, interquartile range; PLA2R, phospholipase A2 receptor; RU, relative units. Differences of median were compared with the Mann-Whitney test. One out of the 36 patients (2.7%) in the corticosteroid-cyclophosphamide group, and 3 out of the 25 patients (12%) in the tacrolimus-rituximab group who had achieved partial remission presented a relapse (Supplementary Table S8). The 3 relapses in the tacrolimus-rituximab group occurred at month 12, 3 months after tacrolimus discontinuation. Two of them occurred in anti-PLA2R–positive patients who had reached an immunologic response at month 6 and were not accompanied by a reappearance of anti-PLA2R antibodies. The remaining relapse occurred in a patient in whom anti-PLA2R antibodies had not been measured at baseline. Tacrolimus was restarted in 2 patients and the other one was treated with rituximab. The relapse in the corticosteroid-cyclophosphamide group occurred at month 9 in an anti-PLA2R–positive patient who had reached immunologic response at month 3. In this case, the relapse was accompanied by a reappearance of anti-PLA2R antibodies and the patient was treated with tacrolimus. All patients except 6 (1 from the corticosteroid-cyclophosphamide group and 5 from the tacrolimus-rituximab group) had at least 1 adverse event (Table 5). Most adverse events were of low (345, 84%) or medium (56, 13%) severity, but 17 (4%) were serious. There were more adverse events and more adverse events per patient in the corticosteroid-cyclophosphamide group than in the tacrolimus-rituximab group (P = 0.04). More patients in the corticosteroid-cyclophosphamide group had leukopenia and Cushing syndrome, while acute kidney injury, hyperkalemia, diarrhea, and distal tremor were more common in the tacrolimus-rituximab group. There was a statistically significant association between the presence of leukopenia and the development of infections, both in the total trial cohort and in each treatment group (P < 0.0001 for the total group and for the cyclophosphamide group, P = 0.041 for the tacrolimus/rituximab group). Most adverse events occurred within the first 9 months of the trial (304 of 409, 74%), but only 5 of the 17 serious adverse events occurred within this period.Table 5Adverse events according to treatment groupEventCorticosteroid-cyclophosphamide (n = 43)Tacrolimus-rituximab (n = 43)P valueaP values are for the differences in number of events between groups.PatientsEventsPatientsEventsAny adverse event42 (98)239 (411)39 (91)170 (280)0.04Serious adverse events8 (19)10 (17)6 (14)7 (12)0.93 Fatal0 (0)0 (0)0 (0)0 (0)1.00 Nonfatal8 (19)10 (17)6 (14)7 (12)0.56Nonserious adverse events34 (79)229 (394)33 (77)163 (268)0.04Adverse events in at least 5% of patients or any serious adverse eventSystemic/general Headache3 (7)5 (9)5 (12)7 (12)0.48 Asthenia7 (16)9 (16)4 (10)4 (7)0.52 Depression3 (7)3 (5)1 (2)2 (3)0.62 Anxiety8 (19)14 (24)4 (10)4 (7)0.35Metabolic Cushing syndrome7 (16)8 (14)0 (0)0 (0)0.01 Hyperglycemia4 (9)4 (7)2 (5)2 (3)0.68 Hyperlipidemia5 (12)7 (12)2 (5)2 (3)0.43Gastrointestinal Diarrhea4 (9)4 (7)13 (31)13 (21)0.02 Abdominal complaint7 (16)7 (12)3 (7)4 (7)0.31Infections Upper respiratory infection10 (23)14 (24)10 (24)14 (23)1.00 Pneumonia4 (9)4 (7)2 (5)2 (3)0.68 Urinary tract infection4 (9)4 (7)1 (2)1 (2)0.36 Tuberculosis1 (2)1 (2)0 (0)0 (0)1.00Musculoskeletal Bone pain10 (23)12 (21)9 (21)12 (20)1.00 Myalgia7 (16)9 (16)2 (5)2 (3)0.16 Cramps5 (12)6 (10)5 (12)5 (8)1.00 Bone fracture1 (2)1 (1)2 (5)2 (3)0.62Neurologic Tremor2 (5)2 (3)7 (17)7 (12)0.09 Paresthesia2 (5)2 (3)2 (5)2 (3)1.00Hematologic Anemia13 (30)17 (29)9 (21)10 (17)0.46 Leukopenia13 (30)22 (38)2 (5)2 (3)0.003Cardiovascular and blood pressure Acute coronary syndrome1 (2)1 (2)1 (2)1 (2)1.00 Venous thrombosis5 (12)5 (9)2 (5)2 (3)0.43 Hypertension5 (12)7 (12)6 (14)6 (10)0.76 Hypotension4 (9)4 (7)0 (0)0 (0)0.12Renal Acute kidney injury8 (19)8 (14)14 (33)16 (26)0.14 Hyperkalemia1 (2)1 (2)6 (14)6 (10)0.06 Edema6 (14)6 (10)4 (10)5 (8)0.74Dermatologic Skin eruption4 (9)4 (7)0 (0)0 (0)0.12Miscellaneous Drug infusion reaction1 (2)1 (2)4 (9)4 (7)0.36 Cancer2 (5)2 (3)1 (2)1 (2)0.62 Gastric adenocarcinoma1 (2)1 (2)0 (0)0 (0) Breast carcinoma1 (2)1 (2)0 (0)0 (0) Rectal carcinoma0 (0)0 (0)1 (2)1 (2)Results for patients are presented as n (%), for events as n (rate per 100 patient-years).a P values are for the differences in number of events between groups. Open table in a new tab Results for patients are presented as n (%), for events as n (rate per 100 patient-years). There were no statistically significan