Primary Mesangial Proliferative Glomerulonephritis: The Forgotten Glomerulopathy

1. Introduction

Mesangial proliferative glomerulonephritis (MesPGN) refers to the light microscopic (LM) finding of mesangial hypercellularity, currently defined as greater than 3 mesangial cell nuclei per mesangial area [1], with or without increased mesangial matrix. Endocapillary hypercellularity, glomerular necrosis, or crescents should not be present. Mesangial proliferation can be focal or diffuse, segmental or global. After utilizing immunofluorescence microscopy (IFM) and electron microscopy (EM), mesangial proliferation may be found in various primary renal diseases, most notably IgA nephropathy (IgAN) [2], or result from systemic illness, such as lupus nephritis, IgA vasculitis with nephritis [3], monoclonal gammopathy of renal significance [4], and post-infectious glomerulonephritis [5]. Other primary glomerulopathies showing mesangial proliferation on LM include fibrillary glomerulonephritis [6] and C3-glomerulopathy [7], but these are readily distinguished by IFM and EM. The topic of this review is primary MesPGN defined as mesangial proliferation not consistent with IgAN or another primary renal disease and unassociated with infection or systemic illness. Notably, in the KGIGO 2021 Guideline on treating glomerular disease [8], a separate section on MesPGN is not included.

MesPGN can was initially highlighted as a separate entity in nephrotic children by White et al. in 1970 [9], although it can occur at any age. Clinically, patients with MesPGN may present as steroid-sensitive (SS) nephrotic syndrome (NS), although relapses are common. However, a significant minority are steroid-resistant (SR) from the outset and may also be unresponsive to additional immunosuppressive agents. Microscopic hematuria is found in most and may occasionally be gross. Other patients present with subnephrotic proteinuria and/or hematuria. Hypertension and reduced GFR are more common with mesangial proliferation than with MCD, as is progression to ESKD.

The relationship of MesPGN to the primary podocytopathies, i.e., MCD and primary focal segmental glomerulosclerosis (FSGS), remains uncertain. In the setting of idiopathic NS (INS), i.e., NS with no secondary cause, some biopsies with mesangial proliferation only show diffuse foot process effacement (FPE) on EM without electron dense deposits (EDDs) and have no immune deposits by IFM; however, IgM immune deposits by IFM with or without complement are common. Segmental sclerosis may occur regardless of immune deposits or EDDs. Border hypothesized nearly 40 years ago that MesPGN is distinct from the primary podocytopathies [10]. According to his hypothesis, MesPGN may initially have only minimal changes (FPE on EM) or mild mesangial abnormalities on LM. As the disease progresses, more marked mesangial proliferation develops and eventuates in segmental sclerosis. Immune deposits are integrally involved, especially IgM and less commonly IgG, along with complement. Podocytes would be secondarily injured, as opposed to the primary podocytopathies where the podocyte is the primary target. Others consider MesPGN to be part of the primary podocytopathies, perhaps a transition state from MCD to FSGS with supporting evidence coming from experimental animals [11] and human studies [12].

As the focus of this review, we will define the prognostic and therapeutic implications of finding mesangial proliferation in biopsies of patients with INS. We will consider both the significance of segmental sclerosis in primary MesPGN, as well as the significance of mesangial proliferation in primary FSGS. The significance of immune deposits, i.e., IgM and complement, will be discussed in detail. Our recommendations regarding therapy will be presented.

2. Mesangial Proliferation

Mesangial proliferation has been addressed in the pediatric literature for decades [9,13]. The International Study of Kidney Disease in Children (ISKDC) distinguished the adverse outcome of segmental sclerosis on LM compared to those with only minimal changes [9,13]. The significance of abnormal mesangial proliferation was noted, and such cases were labeled MesPGN.

The ISKDC evaluated 521 children with INS and, after excluding FSGS, found by LM that 219 had completely normal glomeruli, 98 had focal global sclerosis, 29 had focal tubular changes, 16 had mild mesangial thickening, 27 had mild mesangial hypercellularity (defined as 3 cells per peripheral mesangial area), and 12 had diffuse mesangial hypercellularity (> 4 cells peripheral mesangial area, the current definition of abnormal mesangial proliferation) [14]. The latter group had a significantly higher percentage of initial steroid non-responders (45.5%) than the others (4.7%– 14.8%), although by 1 year there was no significant differences.

Similarly, the Southwest Pediatric Nephrology Study Group evaluated 29 children with INS and underlying diffuse mesangial hypercellularity without segmental sclerosis [15]. Microscopic hematuria was found in 89%, and 7 had reduced creatinine clearance. Three were IgM positive and 4 contained EDDs. All patients had diffuse FPE (70 – 80%). Of 4 re-biopsied, 1 had FSGS. Overall, 12 of 21 steroid-treated patients had complete remissions and 3 had partial remissions. All 4 with 3+ mesangial hypercellularity (> 5 nuclei per mesangial area) were unresponsive, although 2 subsequently had spontaneous remissions. None of 6 steroid-resistant patients responded to alkylating agents.

Alexopoulos et al. studied 28 adult Greek patients with INS and MesPGN, excluding segmental sclerosis, 25 of which received steroids [16]. Initially, about one-third had complete remission, one-third partial remission, and one-third were unresponsive. At a mean of 64 months, 40% were in complete remission, 36% in partial remission, and 24% remained unresponsive. Although there was no control group, these results are inferior to the reported results for treating MCD, again supporting the tenet that mesangial proliferation is an adverse finding versus MCD. IgM deposits were present in all nonresponders, although they were also present in about 20% with complete remission. Subsequent biopsies in 2 nonresponders showed FSGS. Altogether, these studies highlight the adverse effect of mesangial proliferation in both children and adults with INS in the absence of segmental sclerosis.

In a patient with INS and MesPGN on initial biopsy, is the presence of segmental sclerosis relevant? Waldherr et al. evaluated 38 nephrotic children with diffuse mesangial proliferation on biopsy divided into 18 with no segmental sclerosis (group 1) and 20 with associated segmental sclerosis (group 2) [17]. Clinical presentation was similar, although 2/16 group 1 patients responded to steroids and 2 spontaneously remitted compared to none of the group 2 patients. At final evaluation, 5/18 in group 1 and 10/20 in group 2 progressed to ESKD or had impaired function. Of 11 repeat biopsies in group 1, 8 developed segmental sclerosis (FSGS). Interestingly, 3 of these 8 no longer had mesangial proliferation. Overall, most patients (>95%) had hematuria and 70% had segmental sclerosis initially or on repeat biopsy. Hence, segmental sclerosis in the setting of MesPGN portends a worse prognosis.

Does mesangial proliferation portend a worse prognosis if found on an initial biopsy showing FSGS? The evidence is conflicting. Schoeneman et al. assessed the significance of “marked mesangial accentuation or proliferation” in 24 children with steroid-unresponsive NS diagnosed as FSGS [18]. Ten of the 13 with mesangial involvement had reduced glomerular filtration rates (< 90 ml/min/1.73m²) at the end of follow-up (4 reaching ESKD) compared to none of 11 without mesangial involvement. Neither group was steroid-responsive. Studying 80 nephrotic adults with FSGS, Ponticelli et al. found that only mesangial proliferation and the lack of remission (complete or partial) with immunosuppression significantly predicted the risk of dying or doubling of serum creatinine [19].

By contrast, Nickavar and Rahbar evaluated 46 pediatric nephrotic FSGS patients and found mesangial hypercellularity did not correlate with steroid response, although there were significant correlations with anemia and hypertension [20]. Schwartz et al. studied 81 adult FSGS patients and found that 16% had mesangial hypercellularity that did not significantly correlate with ESKD [21].

Hence, it remains uncertain whether the presence or absence of mesangial hypercellularity in the setting of INS with segmental sclerosis (FSGS) is of major relevance as it is in biopsies with MCD where it portends a worse prognosis. In our opinion, in a patient with INS, segmental sclerosis is the dominant prognostic pathologic finding, whether present on initial biopsy or only on subsequent biopsy, regardless of the original biopsy being MCD or MesPGN.

3. IgM Nephropathy

In patients presenting with INS, LM may show MCD, MesPGN, or FSGS, all with diffuse FPE on EM, although MesPGN may have lesser degrees of FPE. Granular mesangial IgM deposits may be found complicating any of the 3 LM appearances. Rarely, crescents may be observed. These deposits may be focal or diffuse, global or segmental; however, IgM restricted only to segmental scars does not apply. Complement components (C3 and/or C1q) may be present, although C1q staining > 2+ defines C1q nephropathy (C1qN, vide infra) regardless of IgM. Other immunoglobulin classes may be present, but IgM must be dominant, especially if IgA is present. Whereas IgM-positive biopsies often have associated mesangial proliferation, well defined deposits may exist with only minimal changes [22]. Mesangial EDDs may or may not be detectable by EM, but significant subendothelial or subepithelial capillary wall deposits should not be found.

Some argue that biopsies containing IgM represent a separate entity, IgM nephropathy (IgMN), versus classification based purely on LM, i.e., MCD, MespGN, or FSGS. Unfortunately, there is no consensus definition of IgMN. Some authors accept 1+ IgM staining as a minimum threshold, others require > 2+. Some require mesangial EDDs on EM. Some authors accept segmental sclerosis, however, cases with IgM restricted to segmental scars are not included. It remains uncertain whether IgMN deserves a separate designation as opposed to the standard classification based on LM alone.

IgM deposits with focal or diffuse mesangial proliferation were first described in 1974 in patients presenting with recurrent macroscopic hematuria or persisting microscopic hematuria +/- low-level proteinuria [23]. C3 and C1q were present in all patients; however, IgA was also present in two-thirds. None developed impaired kidney function. In 1978, two groups simultaneously reported MesPGN with IgM deposits underlying INS or asymptomatic non-nephrotic proteinuria [24,25]. These seminal papers highlight the variable clinical presentation of patients with mesangial IgM depositions, ranging from hematuria with or without non-nephrotic proteinuria to INS with or without hematuria.

4. IgM Deposits: Pathogenicity and Prognostic Relevance

Natural IgM antibodies are polyreactive and germline-derived, produced without prior antigen exposure by B-1 cells in the bone marrow and spleen of humans and in the peritoneum of mice [26]. Besides reacting with pathogens as the first line of defense, natural IgM is important for clearance of apoptotic cells. However, these antibodies may react with self-epitopes expressed on injured tissue, including the kidney, and may activate complement to further injury. Relevant epitopes reactive with these natural IgM antibodies may include cardiolipin and other phospholipids.

Strassheim et al. evaluated the role of natural IgM to further glomerular injury in a murine model of FSGS wherein glomerular IgM, C3, and C4d are readily detectable [27]. Treatment with an anti-CD20 antibody reduced glomerular IgM deposition and significantly attenuated clinical disease. Also, B-cell deficient mice had absent glomerular IgM deposition and significantly less albuminuria, FPE, and glomerular C3 deposition than control mice.

Panzer et al. evaluated the effect of natural IgM to cause and/or contribute to glomerular disease in a non-immune-complex, non-sclerotic mouse model of glomerulopathy, thrombotic microangiopathy from Factor-H deficiency [28]. IgM was deposited in the mesangium and subendothelially, and mice lacking both Factor H and IgM had milder disease. Naturally occurring monoclonal IgM from some clones increased urine albumin in FactorH/IgM deficient mice but not in wild type controls.

Trachtman et al. found deposits of IgM and activated complement fragments (C4d, iC3b/C3d, and human C9 neoepitope) in some glomeruli of 10 patients with FSGS (6 primary FSGS, 4 secondary) [29]. Patients with primary podocytopathies (MCD or FSGS) had elevated plasma levels of C4a and soluble C5b-9. Elevated IgM levels reactive with glomerular endothelial cells were detected by both cell-based assay and ELISA targeting cardiolipin in nephrotic patients. Only manipulated glomerular endothelial cells bound a natural IgM monoclonal antibody as well as IgM from normal serum. The fact that these antibodies only reacted with injured cells suggests that they are not the initiators of injury but may be exacerbators binding to neoepitopes exposed by prior injury.

Altogether, the data indicate a potential pathogenic role for IgM, especially naturally occurring IgM, as a contributor to the pathogenesis of INS, regardless of LM. Whereas the IgM may not initiate injury, it may amplify it. However, evaluating the prognostic significance of mesangial IgM deposition in the setting of INS has produced conflicting results. There is marked heterogeneity in populations studied (pediatric versus adult versus mixed), clinical presentations (INS, non-nephrotic proteinuria +/- hematuria), LM findings (MCD, MesPGN, FSGS, or mixtures), EM findings (EDDs required or not, degree of FPE), and definitions of IgM positivity (> 1+ versus > 2+ required). Clinical criteria are variable (e.g., INS required or series based solely on pathology). Furthermore, in series based on pathology, the specific biopsy practices of the center may influence the frequency with which less severe clinical manifestations are found, such as non-nephrotic proteinuria and/or hematuria. We will discuss selected studies based on the LM findings and age of presentation (pediatric versus adult), highlighting the inconsistent results that hamper making specific conclusions.

5. Childhood IgM Nephropathy

Studies evaluating the significance of IgM deposits in pediatric MCD have been conflicting (see Table 1). Biopsies are typically performed in this population for steroid-dependence (SD) or steroid-resistance (SR) unless patients are < 1 year old or > 10-12 years old. Pardo et al. evaluated 61 children with MCD (mild mesangial proliferation accepted) and found 36 biopsies had mesangial IgM deposits (24 diffuse/global, 12 focal/segmental) with C3 in 11 [30]. There was no significant difference in therapeutic response based on IgM. Al-Eisa et al. evaluated 27 pediatric patients with diffuse mesangial IgM staining, excluding patients with segmental sclerosis, and compared them to 63 diagnosed as MCD, although mesangial expansion was allowed in either group [31]. There was no difference in baseline characteristics, response to steroids, or outcome between the groups.

Vintar Spreitzer et al. studied 55 children with MCD divided into IFM negative cases (29/55), IgM-positive cases (19/55), and C1qN (vide infra) defined as C1q > 2+ intensity regardless of other immunoglobulin staining (7/55) [32]. Diffuse mesangial hypercellularity was present in about 10-15% of each group. EDDs were found in 5/13 with IgM deposits and 6/6 with C1qN. There was no difference between these groups in baseline clinical characteristics, clinical course, or kidney outcome after a median follow-up of 16.9 years.

By contrast, others found adverse features associated with IgM deposition in pediatric MCD. Zeis et al. evaluated 85 nephrotic children with MCD (allowing mesangial proliferation) and found 20 had positive IFM, mainly with IgM, and 65 were negative [33]. On follow-up biopsies, 6/20 (30%) IFM-positive patients evolved into FSGS compared to 3/65 (4.6%) IFM-negative patients. Of 331 children biopsied for non-nephrotic proteinuria and/or hematuria, 44 were IgM-positive and 148 were IFM negative; the remainder had IgA-IgG deposits. Overall, 7/44 (16%) IgM-positive patients evolved into FSGS compared to 0/148 IFM-negative cases.

Other pediatric studies combined patients with MCD, MesPGN, or FSGS by LM and assessed the significance of IgM. Habib et al. assessed 222 French children with INS and found 122 with MCD, 10 with MesPGN, and 90 with FSGS [34]. Of the 222, 54 were IgM-positive, including 33/122 with MCD, 2/10 with MesPGN, 19/90 with FSGS. There was no significant difference in response to steroids based on IgM-positivity. Cases with other immune reactants (C1q, IgG, IgA) also did not respond differently compared to IFM-negative cases.

In a study of 45 Lithuanian children with INS, Juozapaite et al. found 18 were IgM-positive (8 MCD, 1 MesPGN, 9 FSGS based on LM) and 27 IgM-negative (16 MCD, 6 MesPGN, 5 FSGS) [35]. After approximately 4 years of follow-up, there was no significant difference in remission, disease activity, or decrease in kidney function.

Albakr et al. studied 15 Saudi children with IgM-positive biopsies, including 9 MesPGN, 1 FSGS, 4 MCD, and 1 uncertain [36]. Clinically, 14/15 had INS, and all but 1 responded to steroids initially. Relapses were common. Most were in complete remission at last follow-up with 1 progressing to ESKD.

By contrast, other studies combining variable LM demonstrated worse outcomes with IgM-positivity. Swartz et al. reviewed biopsies of 170 Texas children and found that 55 had MCD (24/55 IgM-positive), 43 had MesPGN (20/43 IgM-positive), and 72 had FSGS (14/72 IgM-positive). Of 23 with IgM-positive MCD, 3 (13%) developed ESKD, a much higher rate than expected for MCD [37]. Adjuvant therapy for IgM-positive MCD was more effective with cyclosporine (7/8 responding to initial therapy, 5/6 as secondary therapy) compared to cyclophosphamide (2/11 responding).

Mubarak et al. identified 135 pediatric Pakistani patients with INS and IgM-positive biopsies, including 46 MCD, 52 MesPGN, and 37 FSGS (all 37 with mesangial proliferation) [38]. Comparing a subset of 95 IgM-positive patients to 267 IgM-negative MCD patients showed greater hematuria (28 versus 16%) and progression to kidney failure (16 versus 2.5%), although IgM-positive patients were not restricted to MCD on LM as the controls were.

Kanemoto et al. reviewed 70 Japanese children with steroid-dependent or steroid-resistant INS and found 30 were IgM-positive, including 21 of 53 with MCD, 6 of 10 with MesPGN, and 3 of 7 with FSGS [39]. Overall, the IgM-positive group was more likely steroid-resistant (14/30, 46.7%, versus 11/40, 27.5%, p<0.05). Notably, all 14 steroid-resistant IgM-positive cases responded to cyclosporine, mimicking the results of Swartz noted above.

6. Adut IgM Nephropathy

Studies in adults are similar (see Table 2). Yang et al. evaluated 37 adult with INS and MCD by LM divided into 12 with IgM deposits and 25 without [40]. The baseline clinical presentation and response to treatment was not different between the 2 groups with the exception that the recurrence rate was significantly higher with IgM deposition.

Lee et al. evaluated 63 adult MCD patients and found 15 had positive IFM, including 9 with IgM deposits [41]. Other deposits included C1q in 4, IgG in 4, IgA in 3, and C3 in 1. By multivariable analysis, the presence of deposits was the only significant factor predicting the composite endpoint (doubling of serum creatinine, ESKD, or death), which occurred in 4/15 (26.7%).

Little et al. evaluated 54 predominantly adult patients with MesPGN by LM and diffuse IgM deposits, including 22 with INS, 14 with non-nephrotic proteinuria, with (54%) or without (46%) hematuria, and 18 with isolated hematuria (gross in 10) without proteinuria [42]. Four of the 22 INS patients progressed to ESKD, 2 of whom were re-biopsied and had segmental sclerosis. Of 20 patients with INS given steroids, 12 had remission, 10 of whom became steroid-dependent. Of 11 receiving cyclophosphamide, 3 had lasting response, 6 relapsed upon stopping therapy, and 2 had no response. Of 10 given cyclosporine, 2 had a lasting response, 5 relapsed after remitting, and 3 had no response. None of the patients with non-nephrotic proteinuria or isolated hematuria were immunosuppressed and none had progressive loss of kidney function over 7 years. There was no control group having MesPGN without IgM.

Connor et al. evaluated 57 British adult cases with predominant mesangial IgM staining accompanied by mesangial EDDs in all cases [43]. By LM, 13 (23%) had MCD and 40 had FSGS (70%); mesangial proliferation alone or with FSGS was found 17 (30%). Less than half (22/57, 38.6%) had nephrotic syndrome, 26 (45.6%) had non-nephrotic proteinuria, 6 presented with hematuria, and 3 had declining kidney function at the time of biopsy. Of 23 nephrotic patients, 14 achieved partial remission with 9 achieving complete remission. Doubling of serum creatinine occurred in 25% at 3 years and 30% at 5 years. Seven reached ESKD with 6/7 having segmental sclerosis on > 20% of glomeruli as the worst histologic prognostic feature. There was no adverse effect of mesangial proliferation. Ten patients were re-biopsied, with 8 showing FSGS.

Chae et al. evaluated 94 adult Korean patients with IgMN, including 25 MCD, 48 MesPGN, and 21 FSGS [44]. All 25 with MCD had INS with urine protein/creatinine ratio (UPCR) of 6.3, but proteinuria levels were significantly lower with MesPGN (UPCR of 0.4, p<0.001) and trended less with FSGS (UPCR 2.1, p=0.089). However, the LM subtype predicted neither eGFR nor proteinuria evolution over time. When compared to IgM-negative patients with MCD, MesPGN, or FSGS, the IgM-positive group as a whole and the 3 IgM-negative subgroups all fared the same.

Yun et al. evaluated 63 adult patients with IgMN defined as dominant mesangial IgM including cases with segmental sclerosis [45]. Of 63 patients, 13 had EDDs in addition to IgM with no significant difference in outcome compared to the 50 without EDDs. The full 63 were then compared to 469 with IgAN, 103 with IgM-negative FSGS, and 91 with IgM-negative MCD. The IgMN group had a similar percentage of glomeruli with segmental sclerosis (6.4%) as the FSGS patients (8.7%, p=0.13), and a similar percentage reached the primary adverse renal outcome of > 50% decrease in eGFR, eGFR < 15, or ESKD (46% for IgMN versus 41% for FSGS, p=0.52). These data indicate that in the presence of segmental sclerosis, the presence of IgM did not impact prognosis. IgMN patients overall fared significantly worse than MCD patients without IgM deposits (46% versus 19% reaching renal end-point, p < 0.001); however, the apparent adverse effect of IgM deposits with MCD compared to MCD alone could not be assessed since IgMN included segmental sclerosis.

Combining children and adults, the presence of diffuse mesangial IgM-positivity is an adverse prognostic factor in those without segmental sclerosis on initial biopsy. The progression to segmental sclerosis on repeat biopsies in those with MCD or MesPGN initially appears more frequent with IgM-positivity. Whether IgM-positivity imparts worse prognosis in the presence of segmental sclerosis is less certain.

7. IgM in Segmental Scars

The presence of IgM +/- C3 restricted to segmental scars does not define IgMN and has been considered of no pathogenic significance. However, Zhang et al. evaluated 106 adult patients with primary FSGS and found that 58 (54.7%) had IgM deposits mainly restricted to segmental scars, a minority of which were C3-positive [46]. None of the IgM-negative patients were complement-positive. Only patients with dual positivity (IgM and C3) were at higher risk for refractory NS (defined as non-remitting or SD), with odds ratio 11.32 (p=0.003); those with IgM alone were not. Regarding renal dysfunction, the presence of IgM significantly increased risk regardless of C3 deposition. Importantly, IgM-positivity was mainly restricted to scars and was not present diffusely in the mesangium, suggesting that IgM and complement may be amplifiers of disease in FSGS even when restricted to segmental scars.

Similarly, Mirioglu et al. evaluated 86 Turkish adult primary FSGS patients for the IgM +/- C3 deposits and found 22 had both (dual positivity), 22 had only IgM and 42 had neither. However, IgM positivity could include deposits restricted to scars or involving unaffected mesangium. Patients with dual positivity had higher proteinuria, lower eGFR, and a greater degree of segmental sclerosis compared to the other 2 groups which did not differ. Dual positive patients more commonly reached the primary endpoint (> 50% decline in eGFR or ESKD, hazard ratio 3.36, p=0.009); they also had lower complete/partial remissions.

By contrast, Pacic et al. studied adult patients with primary FSGS and found no significant relationship between dual IgM and C3 positivity with a composite of > 50% decline in eGFR, ESKD, or death, although IgM did impart a worse prognosis in secondary FSGS cases [47]. Peng et al. studied 264 Chinese primary FSGS patients and found neither C3 nor IgM deposition considered separately predicted an adverse renal outcome (> 50% reduction of eGFR, ESKD, death from kidney disease), although C3 deposits in the capillary wall in addition to mesangium did [48]. IgM and C3 together were not considered. Looking at 175 patients with primary FSGS, the CureGn Consortium found neither C3 nor IgM staining predicted the composite outcome (> 40% decline in eGFR or ESKD) by multivariable analysis, but complement activation as assessed by urinary soluble C5b-9 did, mimicking work of prior authors [49]. Such complement activation further implicates an active pathogenic role for IgM, although both animal [50,51] and human studies [49,52] of FSGS also indicate a role for alternate pathway of complement activation in addition to IgM-mediated classical pathway activation.

Overall, these data support a pathogenic role for IgM in the primary podocytopathies, including MesPGN presumably through activation of complement via the classical pathway. In IgM-positive MesPGN, one or more undefined mesangial antigens may be the primary target of IgM antibodies initiating the disease with secondary podocyte injury. IgM may also be secondarily involved as an amplifier of disease by reacting to neoepitopes exposed by mesangial injury initiated by other mechanisms. It remains uncertain to what degree any detectable IgM is composed of naturally occurring, low-affinity, polyreactive IgM versus T-cell dependent, high-affinity IgM. Given a pathogenic role for IgM, B-cell depletion (e.g. rituximab) seems a reasonable therapeutic option in positive cases.

8. C1q Nephropathy

First describe in 1985 in older children and young adults with proteinuria and hematuria, C1qN was characterized by bright C1q immunofluorescence with variable LM, including MCD, MesPGN, proliferative glomerulonephritis, and FSGS (see Table 3) [53]. However, on IFM, a “full-house” pattern of positive staining (IgG, IgM, IgA, C3, C1q) was found, mimicking systemic lupus erythematosus (SLE) nephritis with the only difference being the lack of endothelial cell tubuloreticular inclusions (TRI) on EM; EDDs were predominantly mesanagial/paramesangial with scant capillary wall deposits. No patient had symptoms/signs/serologies supporting SLE, and complement levels were normal. The biopsies were not consistent with immune complex-membranoproliferative glomerulonephritis (IC-MPGN), the other glomerulopathy likely to stain for C1q. The response to steroids was poor. Iskandar et al. reported 15 pediatric cases with nearly identical clinical and pathologic features [54].

In 2003, Markowitz et al. found 19 (0.21% of biopsies) children and adults with C1qN defined as dominant/co-dominant mesangial C1q staining, mesangial EDDs, and no evidence for SLE [55]. IgG was present in all cases but only 4/19 (21%) had “full-house” positivity. About 80% had nephrotic-range proteinuria and 50% had INS; 22% had hematuria. By LM 17/19 had FSGS, 1 MCD, and 1 MesPGN. Response to steroids was poor. Hence, they posited that C1qN belongs in the primary podocytopathy disease spectrum.

Sharman et al. identified 9 British patients initially considered to have seronegative lupus nephritis otherwise consistent with C1qN [56]. All 9 had “full-house” positivity with negative SLE serology. All 9 had mesangial deposits, but unlike other cases of C1qN, all 9 had subendothelial and 3 had subepithelial EDDs; however, none had the endothelial TRIs typical of lupus nephritis. None developed SLE clinically or serologically after median 6-year follow-up. There was no consistent response to immunosuppression.

Fukuma et al. reviewed 30 Japanese children with C1qN, including 18 with asymptomatic urinary sediment abnormalities detected on routine screening and 12 with INS [57]. By LM, 22 had MCD, 6 had MesPGN, and 2 had FSGS. All 12 with INS received prednisone +/- cyclosporine; at follow-up, 3 were in remission, 8 were frequent relapsers, and 1 reached ESKD. Of the asymptomatic 18, 8 were normal, 9 had persisting urinary sediment abnormalities, and 1 reached ESKD.

Expanding on this work, Hisano et al. added 31 more patients to give 61 total Japanese C1qN patients defined as mesangial C1q > 2+ by IFM, mesangial/paramesangial EDDs on EM (required), and lack of clinical/serologic evidence of SLE [58]. The 61 were divided into 36 with symptomatic urinary sediment abnormalities and 25 with INS. By LM, 46 had MCD, 7 had MesPGN, and 8 had FSGS. On IFM, 38/61 (62%) were IgG+, 10/61 were IgM+ (16.4%), and 33/61 (54%) were C3+. All 25 with INS received prednisone +/- cyclosporine; at follow-up, 8 patients had normal urinalyses, but 13 were frequent relapsers. Of 8 patients with repeat biopsies, 3 lost C1q deposits, but 2 of these 3 showed FSGS. One of 25 with INS reached ESKD as did 1/36 with sediment abnormalities.

Kersnik Levart et al. described 12 pediatric C1qN patients from Slovenia, including 6 FSGS, 4 MCD, and 2 MesPGN [59]. Eight presented with INS and were treated with steroids: 1 responded, 4 became steroid-dependent, and 3 were steroid-resistant. Expanding on this work, Vizjak et al. evaluated 72 patients from Slovania, 28 children (including the 12 described above) and 54 adults, with C1qN defined as dominant or co-dominant mesangial C1q > 2+ without evidence for SLE or IC-MPGN [60]. LM revealed MCD 27, MesPGN 20, FSGS 11, or various other lesions. Clinically, 33 presented with NRP/NS, including 17/27 with MCD, 5/20 with MesPGN, and all 11 with FSGS. Others presented with non-nephrotic proteinuria or hematuria. Other reactants included IgG (66.7%), IgM (80.6%), IgA (47%), and C3 (83%); “full-house” was found in 22 (30.6%) predominantly in those with MesPGN by LM. Mesangial EDDs were found in 48 of 53, with 14 also having capillary wall deposits; FPE was > 50% in half of 16 MCD with INS, 12% of 17 MesPGN with INS, and 71% of 7 FSGS with INS. Repeat biopsies revealed progression from MCD to MesPGN in 2 cases and to FSGS in 1 case. They noted 2 main clinicopathologic subsets: one with INS (a frequently relapsing course) and a podocytopathy (MCD or FSGS) by LM, and the other an immune-complex type of glomerulopathy with urinary sediment abnormalities and/or reduced GFR having either MCD or MesPGN by LM.

Wong et al. studied 9 C1qN children from New Mexico with INS [61]. All had MCD by LM. IgG was positive in 100%, IgM 78%, C3 44%. All 9 eventually achieved remission; 4 were initially steroid-sensitive, 3 steroid-dependent, 2 steroid-resistant. All eventually received calcineurin inhibition or mycophenolate. Compared to non-C1qN-MCD patients, C1qN-MCD patients were more frequently relapsing and steroid resistant.

Gunasekara et al. followed 35 Sri Lankan children with C1qN, including 19 with MCD, 3 with MesPGN, and 13 with FSGS/focal global glomerulosclerosis (FGGS) [62]. Thirty-one presented with INS and 4 with proteinuria +/- hematuria. Besides C1q, 27 stained positive for IgM, but other immunoglobulins and complement staining were not reported. Of 13 with MCD treated with steroids having 1 year of follow-up, 11 had complete remission. Compared to C1q-negative MCD patients, the C1q-positive group had more and earlier relapses but otherwise no difference during follow-up. Also, 2/9 treated FSGS/FGGS patients followed for 1 year had complete remission, as did 0/3 with MesPGN.

Peng et al. evaluated 389 Chinese children with primary FSGS and found 18 (4.63%) had C1qN [48]; 10/18 were C3+, 4/18 IgG+, 17/18 IgM+, and 7/18 IgA+ (2/18 were “full-house” positive). All 18 had mesangial EDDs, and 2 had subendothelial deposits, one of which also had subepithelial deposits. Clinically, 7/18 had INS, 13/18 had hematuria, and 5/18 had hypertension. All 18 with INS received steroids and 17 received other immunosuppressive agents. Initially, 5 were steroid-sensitive and 13 steroid-resistant. Compared to 18 age, sex, and period of biopsy-matched C1q-negative primary FSGS patients, there was no significant difference in long-term remission, ESKD, or the combined end-point (> 40% decline in eGFR or ESKD), suggesting that with segmental sclerosis, C1q-posotivity didn’t matter like IgM-positivity.

Kim et al. evaluated 23 adult Korean patients with C1qN defined as > 2+ mesangial C1q-positivity and mesangial/para-mesangial EDDs and compared them to 23 MCD and 23 FSGS patients matched for age, sex, and period-of-biopsy. Of the 23 C1q-positive, 3 had MCD, 1 had increased mesangial matrix, 8 had MesPGN, and 11 had FSGS, including 8/23 with UPCR >3.0. Four developed ESKD. The presence of mesangial hypercellularity, segmental sclerosis, or the degree of FPE was not related to ESKD within the C1q-positive group. However, comparing the C1q-positive group as a whole to the FSGS control group, ESKD was not significantly different (17.4% versus 30.4%, p=0.491); no MCD controls reached ESKD.

To summarize, C1qN is characterized by significant (> 2+) mesangial C1q staining. Other immunoglobulins (IgM, IgG, and/or IgA) are usually present and may stain at the same, lower, or even higher intensity than C1q. A significant minority will have “full-house” positivity (IgG, IgM, IgA, C1q, C3) resembling lupus nephritis; however, TRIs are absent, they are serologically unremarkable, and SLE does not develop over time. IC-MPGN must also be excluded. Like IgMN, the LM of C1qN is variable and includes MCD, MesPGN, and FSGS. In the absence of segmental sclerosis, C1qN has a prognosis worse than C1q-negative MCD; however, C1qN with segmental sclerosis is probably not worse than non-C1qN FSGS.

9. Pathogenesis of Primary Mesangial Proliferative Glomerulonephritis

In patients presenting with hematuria and/or non-nephrotic proteinuria, the mesangium is the likely primary site of disease, especially if the biopsy is consistent with IgMN or C1qN. In the setting of INS with obvious podocyte injury (diffuse FPE) plus mesangial proliferation, however, two main possibilities exist for the initiation of disease: a primary podocytopathy with secondary mesangial activation versus a primary mesangial insult with secondary podocyte dysfunction.

Some authors hypothesize that mesangial proliferation represents a transition state from MCD to FSGS [12,63]. In this scenario, the primary site of disease is the podocyte with mesangial cells secondarily involved. In our opinion, the primary podocytopathies, MCD and FSGS, represent neither two separate entities nor the continuum of a single entity. The diffuse podocyte injury may be the final common phenotype resulting from one or more circulating substances toxic to podocytes and/or autoantibodies targeting podocyte expressed proteins, most notably nephrin, but including others, enhanced by a susceptible genetic background. We recently reviewed in detail these potential circulating factors and autoantibodies [64]. In this scenario, mesangial cells would be secondarily activated to proliferate and produce increased matrix in response to podocyte injury, justifying classification of MesPGN with primary podocytopathies. This would best apply to patients with INS having diffuse FPE and lacking immune complexes on IFM and EDDs on EM.

It is well known that crosstalk between podocytes and mesangial cells exists [65] and may explain how mesangial cells get secondarily activated in primary podocytopathies. Zhou et al. demonstrated that podocyte sonic hedgehog (Shh) is upregulated in animal models of chronic kidney disease and selectively induces mesangial cell activation and proliferation in vitro and in vivo [66]. Subsequently, Liu et al. demonstrated increased production of extracellular vesicles (EV) rich in Shh in injured mouse podocytes that induced activation and proliferation of mouse mesangial cells in vitro [67]. Wu et al. found expression of the C-X-C chemokine receptor type 4 (CXCR4) and angiotensin II were increased and colocalized in injured podocytes along with upregulation of their respective receptors (stromal cell derived factor 1α and the type-1 receptor of angiotensin II) in mesangial cells, highlighting another mechanism of podocyte-mesangial cell cross talk. [68].

Most recently, He et al. demonstrated podocyte upregulation of fibrillin-1 (FBN1) both in patients with proteinuric kidney disease and animal models of glomerular disease. Mesangial cells were activated and proliferated in response to FBN1 both in vitro and in glomerular mini-organ cultures [69]. Podocyte specific knockout of the FBN1gene reduced mesangial cell activation and glomerulosclerosis in an animal model of FSGS. Hence, mechanisms are being unraveled as to how primary podocyte injury can lead to mesangial injury/proliferation, matrix production, and eventual glomerulosclerosis, supporting classifying MesPGN with the primary podocytopathies. Although this scenario would most apply to MesPGN lacking immune-complexes, IgM deposits may represent natural IgM antibodies reacting with injured mesangial cells functioning as secondary amplifiers and not the initiators of the process.

By contrast, MesPGN could result from primary mesangial cell injury with secondary disruption of podocyte function. IgM deposits +/- complement (IgMN) or C1qN would best fit this scenario especially if only segmental FPE. Cross talk between mesangial cells and podocytes is well described even in common kidney diseases such as diabetes [70] and hypertension [71].

Patients with two other immune-complex diseases (IgAN and lupus nephritis) that result in mesangial proliferation with mesangial immune-complexes may clinically develop florid nephrotic syndrome with diffuse FPE in the absence of capillary wall deposits. A significant minority of patients with IgAN present as INS, and on biopsy they can have normal glomeruli or mild mesangial proliferation but diffuse FPE [2]. IgA deposits and EDDs are restricted to the mesangium. Clinically, besides explosive onset of nephrotic syndrome, these cases usually have normal GFR, normotension, and marked steroid sensitivity, mimicking the typical MCD patient.

These cases could represent the chance occurrence of 2 glomerulopathies (MCD and IgAN) [2]. Also, the patient may simply have MCD with lanthanic IgA deposits as may occur in the general population. Most relevant to this discussion, the disease may simply be IgAN with pronounced podocyte injury due to mesangial-podocyte cross-talk. Cytokines produced by activated mesangial cells in IgAN may directly activate podocytes [50], including TNF-α, angiotensin II, TGF-β, and platelet activating factor, and one could postulate the same occurs in MesPGN.

Furthermore, the pathogenesis of IgAN involves deposition of galactose-deficient IgA1-containing immune complexes in the mesangium with mesangial cell activation, analogous to what may occur in IgMN, although the antigenic target in IgMN is undefined. A prime determinant of prognosis in IgAN is the degree of proteinuria, heralding secondary podocyte injury and dysfunction [72], and directly correlating with the degree of podocyte FPE [73]. Effacement is the initial adaptive response of podocytes to insult. With sustained injury, podocyte hypertrophy and then depletion may occur, fostering adhesions to Bowman’s capsule. The same may occur in IgMN, making the not uncommon finding of segmental sclerosis on repeat biopsies of IgMN analogous to the S lesion of the IgAN Oxford classification.

Lupus podocytopathy is an uncommon form of lupus nephritis (~1%) characterized clinically by nephrotic syndrome usually in the setting of active SLE, a steroid-sensitive but relapsing course, and susceptibility to AKI [74,75,76]. Pathologically, minimal changes, mesangial proliferation, or segmental sclerosis may be found with or without mesangial immune complexes. The distinguishing feature of lupus podocytopathy is diffuse FPE (>70%) without the capillary wall deposits (subendothelial or subepithelial) that occur with proliferative and membranous variants of lupus nephritis.

A recent systematic review of lupus podocytopathy found 26 studies involving 259 patients, of which 113 had minimal changes, 91 mesangial proliferation, and 43 FSGS [76]. The vast majority had NRP and most presented at the time of initial SLE presentation. The vast majority responded to steroids, although FSGS cases were less responsive (84.6% vs 98.9% of MCD cases); 20 eventually developed classic lupus nephritis and 3 reached ESKD.

Hence, both experimental and clinical studies support the possibility that primary mesangial disease, mediated by immune complexes or other mechanisms of mesangial cell activation/injury, may result in secondary podocyte injury presenting as INS with FPE in glomerular capillary loops. We suspect that at least some cases of MesPGN presenting as INS are mediated by this mechanism, especially those with positive mesangial immune-complexes (IgMN, C1qN).

10. Treatment of Primary Mesangial Proliferative Glomerulonephritis.

Supportive care is indicated for all patients, including reducing overweight/obesity, exercise, smoking cessation, sodium restriction, strict BP control, RAS inhibition, and SGLT2 inhibition. Sparsentan, a dual blocker of the angiotensin II-type-1 receptor and the endothelin type-A receptor, was recently shown to significantly reduce proteinuria in patients with primary FSGS, although no significant effect on eGFR decline was demonstrated [77]; sparsentan has recently received FDA approval for FSGS without nephrotic syndrome.

The use of immunosuppression is guided by the clinical presentation. In patients with non-nephrotic proteinuria and/or hematuria, immunosuppression is not indicated. In cases with persisting proteinuria and eGFR declining for no other reason, especially in a child or young adult, immunosuppression may be considered.

11. Initial Immunosuppression

In patients with INS, immunosuppression is indicated regardless of LM, IFM, or EM. There are no RCTs addressing immunosuppression for MesPGN, IgMN, or C1qN. The 2021 KDIGO Guidelines for the treatment of glomerular disease does not include sections on any of the 3. Herein we consider treatment recommendations for all 3 under the rubric MesPGN unless otherwise specified. Given the relationship to the primary podocytopathies, initial treatment should be like that recommended by KDIGO for the initial treatment of MCD/FSGS i.e. high-dose prednisone (1 mg/kg/day for 12–16 weeks, with tapering if responsive). Several RCTs compared alternative initial regimens to high-dose steroids (Table 4). None specifically targeted MesPGN, but we suspect trials of MCD and/or FSGS included patients with mesangial proliferation.

Table 5. Randomized Controlled Trials for Treatment of Frequently Relapsing/Steroid Dependent Primary Podocytopathies Potentially Applicable to Mesangial Proliferative Glomerulonephritis with Nephrotic Syndrome.

Trial	Age	Number	Histology	Treatment	Primary Endpoint	Results	Comment
Sinha 2024 [93]	P	160	INS Biopsy N/A	Levamisole (2-2.5 mg/kg on alternate days) versus prednisolone (0.5-0.7 mg/kg on alternate days)	Proportion with frequent relapses	22.5% versus 40% Prednisone was not non-inferior	Similar proportion in sustained remission, relapses comparable
Wang 2025 [94]	P	270	INS Biopsy N/A	Tacrolimus versus MMF	1-year relapse free survival	Relapse-free survival significantly greater with tacrolimus (HR, 2.86, p<0.001)	Tacrolimus group required less steroids
Basu 2018 [97]	P	120	MCD: 85 FSGS: 35	Tacrolimus plus tapering alternate day steroids versus rituximab 375 mg/m2 twice	12-month relapse-free survival	63% versus 90%, p<0.001 in favor of rituximab	Lower steroid exposure with rituximab
Iijima 2014 [98]	P	48	MCD: 44 FSGS: 3	Rituximab 375 mg/m2 weekly x4 versus placebo	Relapse-free period	267 days versus 101 days, p<0.0001 in favor of rituximab	No significant difference in serious adverse events
Isaka 2025 [99]	A	72	MCD: 57 FSGS: 5	Rituximab 375 mg/m2 on weeks 1, 2, 25 versus placebo	Relapse-free at 49 weeks	87.4% versus 38%, p<0.001 in favor of rituximab	Hazard ratio for relapse with rituximab 0.16, p<0.001
FSGS: focal segmental glomerulosclerosis; HR: hazard ratio; INS: idiopathic nephrotic syndrome; MCD: minimal change disease; MMF: mycophenolate mofetil; N/A: not available; P: pediatric;

Table 5. Randomized Controlled Trials for Treatment of Frequently Relapsing/Steroid Dependent Primary Podocytopathies Potentially Applicable to Mesangial Proliferative Glomerulonephritis with Nephrotic Syndrome.

Trial	Age	Number	Histology	Treatment	Primary Endpoint	Results	Comment
Sinha 2024 [93]	P	160	INS Biopsy N/A	Levamisole (2-2.5 mg/kg on alternate days) versus prednisolone (0.5-0.7 mg/kg on alternate days)	Proportion with frequent relapses	22.5% versus 40% Prednisone was not non-inferior	Similar proportion in sustained remission, relapses comparable
Wang 2025 [94]	P	270	INS Biopsy N/A	Tacrolimus versus MMF	1-year relapse free survival	Relapse-free survival significantly greater with tacrolimus (HR, 2.86, p<0.001)	Tacrolimus group required less steroids
Basu 2018 [97]	P	120	MCD: 85 FSGS: 35	Tacrolimus plus tapering alternate day steroids versus rituximab 375 mg/m2 twice	12-month relapse-free survival	63% versus 90%, p<0.001 in favor of rituximab	Lower steroid exposure with rituximab
Iijima 2014 [98]	P	48	MCD: 44 FSGS: 3	Rituximab 375 mg/m2 weekly x4 versus placebo	Relapse-free period	267 days versus 101 days, p<0.0001 in favor of rituximab	No significant difference in serious adverse events
Isaka 2025 [99]	A	72	MCD: 57 FSGS: 5	Rituximab 375 mg/m2 on weeks 1, 2, 25 versus placebo	Relapse-free at 49 weeks	87.4% versus 38%, p<0.001 in favor of rituximab	Hazard ratio for relapse with rituximab 0.16, p<0.001
FSGS: focal segmental glomerulosclerosis; HR: hazard ratio; INS: idiopathic nephrotic syndrome; MCD: minimal change disease; MMF: mycophenolate mofetil; N/A: not available; P: pediatric;

Li et al. randomized 119 adult Chinese MCD patients to tacrolimus monotherapy versus high-dose oral prednisone with both groups initially receiving 10 days of iv methylprednisolone and found tacrolimus non-inferior with no difference in remission at 12 weeks, time to remission, or relapse rate [78]. Medjeral-Thomas et al. randomized 50 adult British MCD patients to tacrolimus monotherapy or high-dose prednisone and found no difference in complete remission at 8 weeks, 16 weeks, or 26 weeks; there was no significant difference in relapse rate or serum creatinine over time [79]. Chin et al. randomized 144 adult Chinese MCD patients to tacrolimus plus low-dose prednisone (0.5 mg/kg) or the standard high-dose prednisone regimen and found no difference in complete remission by 8 weeks or time to remission but significantly less relapses with maintenance tacrolimus [80]. Hence, a tacrolimus-based regimen is an alternative for initial therapy of adults with MesPGN wishing to avoid high-dose steroid exposure. However, it remains uncertain how much concurrent steroid exposure, if any, is required.

Remy et al. randomized 116 French adults with MCD to mycophenolate sodium and low-dose prednisone or the high-dose prednisone regimen with no significant difference in complete remission at 4 weeks, 8 weeks, or 24 weeks [81]. The INTENT investigators randomized 269 German children with their initial episode of INS in remission within 28 days of high-dose steroids to either MMF for 12 weeks plus alternate day prednisone over the first 2 weeks versus full-dose prednisone for 6 weeks plus alternate day prednisone for another 6 weeks [82]. The primary endpoint, treated relapses within the 24 months following the initial 12-week treatment, occurred in 79.1% of the MMF group versus 74.8% of control showing statistical non-inferiority for MMF with significantly fewer steroid-related side-effects.

Kristensen et al. randomized 67 Danish adult MCD patients to active vitamin D (alfacalcidol) plus low-dose prednisone or high-dose prednisone and found no difference in remission at 16 weeks with significantly fewer adverse events and no difference in relapse rates [83].

Although no RCTs assessed rituximab as initial therapy for podocytopathies, observational data from 3 small retrospective studies totaling 25 adult MCD patients showed a benefit [84,85,86]. Li et al. performed a prospective cohort study in 74 adult Chinese MCD patients comparing rituximab 1g twice in 20 patients, half-dose prednisone (0.5 mg/kg) plus rituximab 1 g once in 28, and full-dose prednisone in 26 as initial therapy and found complete remissions in 50%, 96%, and 96%, respectively, at 12 months with reduced steroid-induced diabetes in the half-dose steroid regimen [87]. Rituximab alone appeared less effective. By contrast, Zhao et al. compared rituximab (4 weekly doses of 375 mg/m²) to standard full-dose prednisone in 82 adult Chinese MCD/FSGS patients, including 9 MCD patients receiving rituximab as initial therapy [88]. All 9 attained remissions, including 8 complete.

Wang et al. compared 14 Chinese MCD patients (12 adults) treated with rituximab (various dosages) +/- short-course high-dose prednisone to 14 propensity-matched patients treated with the high-dose regimen and found 100% remissions overall (11 complete remissions in each group) at 24 weeks; 5 rituximab-treated patients with a year or more of follow-up had no relapses, although 3 other patients relapsed after 24-weeks [89].

The results using rituximab as initial therapy for FSGS are less encouraging, although there are minimal data. Roccatello et al gave 8 adult primary FSGS patients high-dose rituximab (8 weekly infusions of 375 mg/m²) and attained only a partial remission in one patient [90].

12. Frequently Relapsing or Steroid Dependent INS

Such patients are steroid-sensitive, so the presence or absence of segmental sclerosis does not change the approach. For children with frequently relapsing/steroid dependent (FR/SD) INS, KDIGO 2025 guidelines for children recommends one of 5 agents: CNIs, cyclophosphamide, levamisole, mycophenolate mofetil (MMF), and rituximab in no special order [91], For adults with FR/SD MCD, the KDIGO 2021 guidelines recommend cyclophosphamide, rituximab, CNIs, or mycophenolic acid analogs as steroid-sparing agents, without preferring one class [92]. For FR/SD FSGS, the recommendations would be the same.

No trials specifically targeted MesPGN. Several RCTs addressed steroid sparing agents in FR/SD INS. Sinha et al. randomized 80 Indian children with FR/SD INS upon attaining remission to oral prednisolone (0.5 – 0.7 mg/kg) on alternate days or daily during infection versus oral levamisole on alternate days for 1 year and found frequent relapses in 40% with prednisolone and 22.5% with levamisole (p=0.0091); prednisolone was not non-inferior [93]. However, the proportion of patients in sustained remission was similar as was the incidence density of relapses.

Wang et al. randomized 270 Chinese children/adolescents with FR/SD INS in remission to tacrolimus or MMF for 1 year and found tacrolimus was superior in preventing relapse (82% versus 58.5%, p<0.001) and showed a significantly lower cumulative steroid dosage [94].

Rituximab has been studied in several RCTs and numerous observational studies for FR/SD INS in both children and adults. Various dosing regimens have been used. High dose regimens are approximately 1,500 mg/m² (4 weekly doses of 375mg/m²or 1 gram twice, 15 days apart). Moderate dosing is approximately 750 mg/m² (two 375 mg/m² weekly doses or a single 1-gram or 750 mg/m² dose). Low dose rituximab is a single 375 m/m² dose. A very low dose of 100 mg/m² was used but associated with a significant increase in relapse rate [95]. Although a minority of patients with FR/SD INS receiving rituximab will not relapse without subsequent rituximab or other immunosuppression [96], the majority will. Factors affecting relapse rate include the initial dosage schedule, concurrent maintenance immunosuppression, or repeated rituximab dosing.

Basu et al. randomized 120 Indian children with SD INS (~70% MCD, 30% FSGS) in remission to moderate dose rituximab (all 60 received 2 doses of 375 mg/m²) or tacrolimus with both groups receiving tapering alternate-day prednisolone. Rituximab had a significantly higher 12-month relapse-free survival (90% versus 63%, p<0.001. The cumulative steroid dose was lower with rituximab (22.5 versus 86.3 mg/kg) [97].

Two placebo-controlled, double-blind RCTs from Japan showed a significant ability of rituximab to reduce the relapse rate. Iijima et al. randomized 48 children/adolescents (44 MCD, 3 FSGS, 1 unknown) with complicated (defined as receiving additional immunosuppressants plus steroids) FR/SD INS in remission to high-dose rituximab or placebo [98]. Patients also received prednisolone 60 mg/m² on alternate days tapered over 6 weeks. The median relapse-free period was significantly longer with rituximab (267 days versus 101 days, p<0.001), and serious adverse events were non-significantly less with rituximab (26% versus 42%). Isaka et al. randomized 66 adults with FR/SD INS (57 MCD, 5 FSGS, 4 other) in remission to moderate-dose rituximab (375 mg/m² on weeks 1, 2 and 25) or placebo [99]. Steroids or other immunosuppressants were weaned similarly in both groups. Rituximab resulted in a significantly greater relapse-free survival at week 49 (87.4% versus 38%, p<0.001).

Observational data support rituximab for FR/SD INS. Ruggenenti et al. gave rituximab 375 mg/m² (with a second dose if B-cells were still detectable at 1 week) to 30 patients (10 children) with SD/FR INS in remission, including 22 with MCD or MesPGN and 8 with FSGS, and found that all were in remission at 1 year, including 18 treatment-free, and 15 never relapsing [100]. Total relapses decreased from 88 the year before rituximab to 22 after, and the per-patient steroid median maintenance dosage decreased from 0.27 mg/kg to 0 mg/kg (P<0.001). A recent systematic review and network meta-analysis found rituximab the most favorable agent for pediatric FR/SD INS with the fewest relapses and resulting in lowest steroid exposure [101].

Maxted et al. retrospectively evaluated 60 pediatric FD/SD INS patients comparing low, medium, and high dose rituximab and found no difference in 12-month release rate between the various dosing strategies [102]. By contrast, Chan et al. in a multinational study, evaluated 511 children with FR/SD INS in remission of which 191, 208, and 112 were treated with low, medium, or high dose rituximab, respectively. Additionally, 283 (55%) received maintenance immunosuppression (corticosteroids, MMF, or CNIs). Relapse-free survival was significantly shorter with the low-dose regimen without maintenance immunosuppression. However, the low-dose regimen with maintenance immunosuppression had similar outcomes to the moderate and high-dose regimens which were not significantly different from each other, with or without maintenance immunosuppression.

In a placebo-controlled, double-blind, pediatric, RCT of Japanese patients with complicated FR/SD INS, 78 were initially given high-dose rituximab and then randomized to MMF or placebo for 505 days; no significant difference in treatment failure by end of follow-up (median 784 versus 472.5 days, p=0.07) was found, although a post-hoc analysis showed significant benefit if analysis was restricted to time on study drug [103].

Ito et al. found MMF effective in reducing relapse in 9 complicated pediatric SD patients following rituximab compared to historical controls receiving just rituximab [104]. In a non-randomized trial in 29 Japanese children with complicated SD INS given low-dose rituximab, Fujinaga et al. found cyclosporine more effective than MMF for maintaining remission following rituximab [105].

Maintenance rituximab dosing to prevent relapses in SD/FR INS in both children and adults has been studied. An international retrospective study evaluated 346 children with FR/SD INS receiving 2 or more courses of rituximab. Relapse-free survival increased progressively with each subsequent course from 10 months to 12 to 16 months despite no difference in duration of B-cell depletion. Side-effects were considered acceptable (hypogammaglobulinemia in 50.9%, neutropenia 3.7%, infection 4.5%). Similarly, Sinha et al. found sequential rituximab doses effective in reducing relapses in a single-center study of 250 Indian children with complicated SD, steroid-resistant-CNI-dependent, or steroid/CNI-resistant INS and found a reduction of 2 relapses/person-year, a reduced prednisone dose, and withdrawal of additional agents in 62%. Hypogammaglobulinemia occurred in 35%.

Several small studies in adults with MCD reported repeated low doses of rituximab (200 mg or 375 mg/m²) given at 6-month intervals effective in preventing relapses [106,107,108]. In an international retrospective study of adult patients (119 MCD, 64 FSGS) with difficult-to-treat INS (68% SD/FR, 22% steroid-resistant, 85% treated with > 2 immunosuppressants) given rituximab at variable dosages, the RITERM Study Team found 151 (82%) in remission at 6 months of which 83 (55%) achieved relapse-free survival at 36-months. Maintenance rituximab, given to 25%, significantly increased the chance of relapse-free survival with withdrawal of all immunosuppressive medications (61% versus 36%, p=0.01) [109]. Similarly, a French, retrospective, multicenter study involving 134 adult patients with podocytopathies given rituximab found relapse frequency was significantly reduced and maintenance dosing between 6 and 12 months significantly reduced relapse risk (hazard ratio, 0.35, p=0.02) [110].

Resistance to rituximab may occur for various reasons, including reduced bioavailability (urinary loss or internal uptake by B-cells) and development of anti-rituximab antibodies (ARA). Rituximab redistributes CD20 into lipid rafts enhancing complement-dependent cytotoxicity (CDC) [111]. Rituximab also promotes apoptosis [112] and antibody dependent cellular cytotoxicity (ADCC) [113]. More potent B-cell depleters have been developed. Ofatumumab, a fully human type 1 anti-CD20 antibody, is more effective than rituximab in fixing complement as it binds to a different epitope, one closer to the cell membrane, facilitating better complement activation [114]. Obinutuzumab is a humanized type 2 anti-CD20 antibody with a glycoengineered Fc region having increased affinity for FcγRIIIa receptors that are expressed on effector cells (NK-cells, macrophages) markedly enhancing ADCC [115]. Rituximab and obinutuzumab recognize overlapping epitopes but bind in different orientations. Since CD20 is not redistributed into lipid rafts, obinutuzumab does not effectively induce CDC [116]. Obinutuzumab also promotes non-apoptotic (lysosomal mediated) programmed cell death [117].

Up to one-third of patients with INS given rituximab may develop ARA [99,118] which may shorten B-cell depletion and/or promote earlier relapses or rituximab resistance. The alternative B-cell depleters ofatumumab and obinutuzumab appear to not cross-react with ARA and have been successfully used in ARA-positive patients with membranous nephropathy [119,120].

An RCT of 140 children and young adults with CNI- and steroid-dependent INS compared ofatumumab to low dose rituximab and found no difference in 12 or 24-month relapse-free survival [121]. A small pediatric study of ofatumumab in multi-drug-resistant INS was stopped for futility [122].

Several studies assessed obinutuzumab for INS. Dossier et al. retrospectively evaluated 41 rituximab-treated French children with FR/SD INS that did not have B-cell depletion (n=8), had shortened B-cell depletion (< 3 months, n=6), or relapsed after B-cell recovery (n=27) given low dose obinutuzumab and obtained B-cell depletion in all lasting a median 8.3 months [123]. A similar median time of B-cell depletion occurred in the 5 with ARA (8.3 months). Remission occurred in 92% at 12 months and 68% at 24 months.

Parmentier et al. administered low-dose obinutuzumab to 32 pediatric ARA-positive French FR/SD INS patients that failed to B-cell deplete with rituximab (n=9), had short B-cell depletion (n=10), or relapsed, and found a median duration of B-cell depletion of 6.5 months with 62.5% relapse-free survival at 24 months; individual ARA concentrations generally declined over time [124]. Zha et al. obtained 100% remission at 6 months in 10 Chinese children with INS and rituximab failure (no remission or relapse within 6 months); B-cell depletion persisted throughout the 6 months [125].

Similar results obtained in adults. Jin et al. treated 6 Chinese MCD patients with steroid- or multidrug-resistant MCD with obinutuzumab and obtained 6 complete remissions with no relapses after approximately 1 year [126]. Lin et al. gave obinutuzumab to 11 Chinese patients with primary podocytopathies (8 MCD, 3 FSGS) with no remission or quick relapse following prior rituximab with or without additional immunosuppression and obtained 10 complete remissions and 1 partial remission [127]. Li et al. gave obinutuzumab to 161 Chinese with SD/FR INS, including 53 with MCD, 28 with MesPGN, and 21 with FSGS [128]. After a median follow-up of 15 months, 148 (91.9%) achieved remission (98% for MCD, 92.9% for MesPGN, and 90.4% for FSGS), including 91 complete remissions. There were 17% relapses among remitters.

13. Steroid Resistant INS

Steroid resistance occurs in 10 – 30% of childhood INS and is typically defined as failure to remit completely with 4 weeks of high-dose steroids [8]. In adults with primary podocytopathies, the resistance rates are somewhat higher (30 – 40%), require longer courses to define (up to 16 weeks), and vary by histologic type (10 – 20% for MCD versus 25 – 65% for FSGS) [8]. Such resistance may occur at the outset of treatment, termed initial steroid-resistance (ISR). By contrast, a patient may respond initially, but upon relapse become steroid-resistant, termed late steroid-resistance (LSR). The pathophysiology may differ. Genetic defects underly a significant minority of ISR cases, perhaps a third of pediatric cases [129], and ISR cases should undergo genetic testing to avoid unnecessary immunosuppression. Non-genetic ISR probably arises from immune dysregulation resulting in circulating factors or autoantibodies injurious to podocytes [64] which may respond to alternative immunosuppression. Anti-nephrin antibodies were found in a minority of children with non-genetic steroid-resistance (14% in one study) and rarely in genetic steroid-resistance (2%) [130]. Ragliante et al. found anti-slit diaphragm IgG using high-resolution microscopy in 27% of ISR cases [131].

By contrast, LSR is not likely genetic, but strongly linked to circulating factors or autoantibodies. In the study of Ragliante et al. cited above, 87% of cases with LSR had anti-slit IgG antibodies [131]. Furthermore, in patients with steroid-resistant podocytopathy progressing to ESKD, post-transplantation recurrence is much more likely with LSR than ISR (92.9% versus 30.2% in one study, OR 30, p<0.001), supporting a circulating factor or autoantibody in LSR [132].

Epigenetic mechanisms may be involved in steroid resistance, including aberrations in DNA methylation, differential expression of micro-RNAs, and histone modifications(73). It is unclear if such mechanisms are more likely involved with non-genetic ISR or LSR. Epigenetic modifications are known to modulate immune cell function and glucocorticoid receptor signaling and theoretically could accumulate over time and with repeated steroid courses, implicating a greater role in LSR. By contrast, a study found SOCS3 promoter hypomethylation in steroid-resistant INS versus steroid-sensitive cases before steroid exposure, and hence potentially contributing to ISR as well [133]. In the future, epigenetic modifications may be the target of personalized therapy, but much more work is required for this nascent field.

Steroid-resistant INS can also be subdivided into cases responding to intensified additional immunosuppression versus those that are multidrug-resistant. Importantly, neither steroid-resistance nor multidrug-resistance is defined by LM.

Regarding treatment for steroid-resistance, KDIGO guidelines recommend CNIs as initial therapy without distinguishing ISR from LSR [8,91]. Several small RCTs involving children and adults with steroid-resistant INS showed a significant benefit for CNI compared to placebo/standard of care [134,135,136] or compared to intravenous cyclophosphamide [137,138]. Meta-analyses [139,140] support CNI as first-line therapy when steroid-resistant. A network meta-analysis of pediatric steroid-resistant INS found CNIs more efficacious than other agents (cyclophosphamide, MMF, leflunomide, chlorambucil); ISR and LSR were not separately analyzed [139]. A recent Cochrane review of RCTs for treatment of pediatric steroid-resistant INS also found CNIs increased the likelihood of remission compared to either placebo/no treatment or cyclophosphamide [141]. No difference between ISR and LSR was evident, although subgroups were small.

In those wishing to avoid CNI toxicity, alternatives are available for steroid-resistant cases. The NIH-funded FSGS Clinical trial found no difference in remission comparing pulse oral dexamethasone plus MMF to cyclosporin in 138 children and adults with primary FSGS [142]. Rituximab has also been studied for steroid- and multidrug-resistant INS. A small pediatric RCT comparing moderate-dose rituximab to standard-of-care involved 31 children with INS resistant to both steroids and CNIs for 6 months; there was no significant proteinuria reduction at 3 months [143]. Multiple very small observational studies assessed the ability of rituximab to induce remission in INS refractory to both steroids and CNIs and found overall remission rates of 18 – 80% (reviewed in [144]).

In the largest observational study to date, an international study of steroid-resistant pediatric INS involving 28 centers, Chan et al. retrospectively analyzed 246 non-remitted patients, including 146 with CNI-resistance (on CNIs for > 6 months) and 100 with CNI administration for < 6 months (technically not yet CNI-resistant) [145]. Included were 112 with ISR and 134 with LSR. Histology included 136 with FSGS, 79 with MCD, 17 with MesPGN, 3 with IgMN, and 1 with C1qN. Overall remission rates at 3, 6, 12, and 24 months ranged from 26% - 39% in the CNI-resistant group and from 42% - 60% in the group on CNIs < 6 months. Both non-remission and partial remission were associated with worse kidney survival compared to complete remission.

Hence, rituximab appears effective in a subset of patients with steroid/CNI-resistant INS, at least in pediatric cases. It would be ideal to identify those most likely to respond. Using high-resolution microscopy, Raglianti et al. found that 12/13 slit-IgG-antibody positive steroid-resistant patients responded to second line immunosuppression versus 2/10 slit antibody-negative. Unfortunately, high resolution microscopy techniques are not routinely available. Perhaps such patients, as well as those with detectable circulating anti-nephrin (and other podocyte antigens) antibodies, may be considered for rituximab therapy. Specifically, for MesPGN, we also feel it is reasonable to consider rituximab for steroid-resistant cases or as first-line therapy in those wishing to avoid high dose steroids if obvious immune-complex involvement is present, i.e., IgMN and C1qN.

14. Discussion

Primary MesPGN is hard to define nosologically, probably because multiple pathophysiological processes can produce that specific histology. If there is diffuse FPE and negative IFM, such cases may be classified with the primary podocytopathies, realizing that MCD, MesPGN, and FSGS are each a group of diseases and not clearly defined separate entities. Circulating factors and/or autoantibodies targeting podocyte antigens are involved in many cases. Mesangial proliferation would be secondary in such cases.

In cases with evidence for immune complex origin (IgMN or C1qN), the mesangium is the likely primary site of disease with secondary podocyte injury. IgMN and C1qN may represent primary immune complex diseases, like IgAN, although the antigenic target(s) remains to be defined. It is uncertain if the immune complexes form in situ, directed against a mesangial antigen, or form in the circulation and deposit, as in IgAN.

It remains uncertain if IgMN deserves to be considered a separate entity, as opposed to just classifying by LM. In the setting of INS, the presence of segmental sclerosis is the dominant prognostic feature, regardless of mesangial proliferation or IgM deposits. The presence of either mesangial proliferation or IgM deposits adds little to prognosis if segmental sclerosis is present but either finding imparts a worse prognosis compared to MCD.

The exact nosological position of C1qN is also unclear. Although bearing many similarities to lupus nephritis, especially when full-house” positivity is present, there are differences. TRIs are absent and patients are negative for ANA, low complement, and anti-double stranded DNA antibodies. Cases rarely evolve into overt SLE. It is possible, although we feel less likely, that C1qN should also be classified with the primary podocytopathies, with the positive immune deposits due to increased mesangial trafficking in the setting of severe proteinuria. Like IgMN, the presence of segmental sclerosis dominates prognosis in C1qN, but compared to MCD, C1qN has a worse prognosis.

It should be stressed that the prognostic implications of mesangial proliferation, segmental sclerosis, or immune deposits apply mainly to the time of biopsy. What ultimately determines the prognosis for the individual patient is the subsequent clinical course. Steroid responsiveness followed by no/infrequent relapses defines an excellent prognosis regardless of histology. Likewise, steroid-resistance, especially if multidrug resistance, markedly increases morbidity and the risk for progressive loss of kidney function, even if MCD.

The approach to treatment of MesPGN lies in the clinical presentation. Non-nephrotic proteinuria and/or hematuria should be treated with general supportive care. Immunosuppression should only be considered if nephrotic syndrome subsequently develops or if GFR declines for no other reason in a patient with significant albeit non-nephrotic proteinuria.

For MesPGN patients (including IFM-negative MesPGN, IgMN, or C1qN) presenting with INS, treatment should mirror that for the primary podocytopathies, as per KDIGO21 and 25 (high-dose steroids, prednisone 1 mg/kg daily). In patients wishing to avoid steroid toxicity from the outset, tacrolimus has the best supporting data. Reduced dose steroids with either MMF or alfacalcidol have support in single RCTs for initial therapy if CNI toxicity is also a concern. There are only limited data on the safety and efficacy of rituximab as initial therapy.

In our opinion, rituximab is the preferred agent for adults with FR/SD INS regardless of LM (MCD, MesPGN, or FSGS) or IFM (IgM, C1q, or negative) using a moderate or high dose regimen. Pediatric guidelines prefer CNIs or MMF for FR/SD INS [146,147]. Hypogammaglobulinemia may be of greater concern in children [148], especially young children [149], where caution is advised by pediatric KDGO 2025 guidelines for those < 7 years-old or receiving multiple courses [91]. If a low-dose rituximab regimen is used, either maintenance immunosuppression (CNI or MMF) should be added. A low maintenance rituximab dose should be considered (e.g., at 6-months or upon B-cell repopulation [150]), especially in those with complicated SD/FR. If there is no remission or early relapse with rituximab, obinutuzumab may be considered, especially if ARA are detected.

Whereas rituximab and other B-cell depleters are reasonably well tolerated, the potential for side effects must be considered [144]. Infusion reactions are not uncommon, occurring in about one-third of patients [151], and include fever, chills, nausea, skin rash, bronchospasm, and hypotension. These can be ameliorated with premedication and a slowed infusion. Serum sickness may occur with repeated infusions in those with ARA [152]. Hypogammaglobulinemia is common (about 10% in children [148] up to 50% or more [149]) and may occur in up to 50% of those given repeated courses [153].Risk factors includeyoung age, low baseline IgG, and longer maintenance immunosuppression [154]. Late-onset neutropenia occurs in < 10% of children [144], although one study found 21% [154]. Infections are not uncommon in children [154] and adults [110], but are usually not severe. Hepatitis B may reactivate following rituximab [155] and prophylaxis in infected patients is recommended. Progressive multifocal leukoencephalopathy has been reported following rituximab but is exceedingly rare and not reported in treatment of INS; however, fulminant viral myocarditis has been reported [156].

For steroid-resistant INS, CNIs have the best data. High-dose dexamethasone plus MMF may be an alternative if CNI toxicity is an issue. Rituximab appears to be effective in a significant minority of cases, although at this time responsiveness cannot be adequately predicted. It remains to be determined if the presence of mesangial immune deposits (IgMN or C1qN), detection of slit diaphragm/podocyte immunoglobulins by high-resolution microcopy, or detection of circulating anti-podocyte antibodies would increase the chance of responding to rituximab or more potent B-cell depleters.

Table 6. Key Takeaways.

Primary mesangial proliferative glomerulonephritis refers to the LM description of > 3 mesangial cells per given area that is not secondary to another primary glomerulopathy (e.g., IgAN, C3G) or systemic illness (e.g., infection, SLE). IFM of primary MesPGN may be negative or contain immune-complexes (IgM +/- complement or C1qN) and EM may or may not contain EDDs Primary MesPGN may present clinically as INS, non-nephrotic proteinuria +/- hematuria, or isolated hematuria Primary MesPGN may be classified with the primary podocytopathies (MCD, FSGS) when presenting as INS, especially in the absence of immune complexes In the setting of immune complexes (IgMN or C1qN), MesPGN may represent primarily mesangial disease analogous to IgAN with secondary podocyte injury Treatment of nephrotic MesPGN is similar regardless if IFM-negative, IgMN, or C1qN Initial IS for nephrosis: high-dose steroids (1 mg/kg/day). Alternatives: racrolimus; possibly half-dose steroids + either MMF or alfacalcidol; rituximab For FR/SD cases: rituximab with maintenance dosings For steroid resistance or multidrug resistamce: CNIs

Table 6. Key Takeaways.

Primary mesangial proliferative glomerulonephritis refers to the LM description of > 3 mesangial cells per given area that is not secondary to another primary glomerulopathy (e.g., IgAN, C3G) or systemic illness (e.g., infection, SLE). IFM of primary MesPGN may be negative or contain immune-complexes (IgM +/- complement or C1qN) and EM may or may not contain EDDs Primary MesPGN may present clinically as INS, non-nephrotic proteinuria +/- hematuria, or isolated hematuria Primary MesPGN may be classified with the primary podocytopathies (MCD, FSGS) when presenting as INS, especially in the absence of immune complexes In the setting of immune complexes (IgMN or C1qN), MesPGN may represent primarily mesangial disease analogous to IgAN with secondary podocyte injury Treatment of nephrotic MesPGN is similar regardless if IFM-negative, IgMN, or C1qN Initial IS for nephrosis: high-dose steroids (1 mg/kg/day). Alternatives: racrolimus; possibly half-dose steroids + either MMF or alfacalcidol; rituximab For FR/SD cases: rituximab with maintenance dosings For steroid resistance or multidrug resistamce: CNIs