Chapter 26. Membranous Nephropathy

• The clinical features of nephrotic syndrome are present.
• The glomeruli are not inflamed.
• The glomerular capillary basement membrane appears thickened.
• There is an absence of glomerular inflammation.
• Idiopathic membranous nephropathy (MN) is relatively common and remains the leading cause of nephrotic syndrome in white adults. It is diagnosed by ruling out secondary causes.
• Secondary MN forms may account for up to one-third of cases and are associated with autoimmune diseases.
• Malignancy increases with increasing age.

Sixty years ago E.T. Bell coined the term membranous glomerulonephritis to describe the renal pathology of a group of patients with the clinical features of nephrotic syndrome in whom the glomeruli were not inflamed but in whom the glomerular capillary basement membrane appeared thickened. The synonymous terms extramembranous nephropathy and epimembranous glomerulonephritis have also been used to describe the disease. The word glomerulonephritis is, however, misleading as one of the main features of the condition is the absence of glomerular inflammation; because of this the word nephropathy is more often employed.

Idiopathic MN is a relatively common immune-mediated glomerular disease and remains the leading cause of nephrotic syndrome in white adults. In the majority of cases, the etiologic agent is unknown, and the disorder is termed idiopathic. Secondary MN forms may account for up to one-third of cases and are associated with autoimmune diseases (eg, systemic lupus erythematosus, SLE), infections (eg, hepatitis B and C), medications [eg, nonsteroidal antiinflammatory drugs (NSAIDs), d-penicillamine, gold], and neoplasias (eg, colon cancer, kidney cancer) (Table 26–1). The association with malignancy increases with age, reaching up to 20% in patients over the age of 60 years. Since idiopathic and secondary forms have similar clinical presentations, the designation of idiopathic is made by ruling out secondary causes by a careful history, physical examination, and laboratory evaluation of the patient. The disease is rare in children, and when it does occur, is commonly associated with an immunologically mediated disorder such as SLE.

There is considerable evidence to support the hypothesis that idiopathic MN is an autoimmune disease. However, the pathogenic mechanisms that cause immune deposits of immunoglobulin G (IgG) and complement to localize predominantly or exclusively on the subepithelial surface of the glomerular basement membrane (GBM) and the subsequent development of proteinuria are not fully understood. The Heymann model of experimental MN in rats, which closely mimics the histologic features of human disease, suggests that the podocyte is the target of injury. The antigenic targets of the antibody response in this experimental model have been localized to the clathrin-coated pits of the podocyte foot processes. They are, specifically, a 515-kDa glycoprotein called megalin, a polyspecific receptor related to the low-density lipoprotein receptor family that functions as a multiligand receptor for the uptake of a variety of macromolecules (eg, aminoglycosides, advanced glycation end products, vitamin D) and a second 44-kDa protein, know as RAP (receptor-associated protein), that is closely related to megalin and is an additional target antigen in the model. Binding of the antibody to the antigen results in activation of the complement with insertion of C5b-9 (membrane attack complex) to the podocyte membrane. Because the antigen–antibody binding and subsequent complement cascade activation occurs on the urinary side of the GBM, and thus away from the circulation, the cellular inflammatory response is blunted, and this accounts for the absence of inflammatory cells in the glomeruli on light microscopy. Although the insertion of C5b-9 to the podocyte membrane is insufficient to cause cell lysis, it does induce podocyte activation and signal transduction, resulting in an increased production of a number of potentially cytotoxic molecules (eg, cytokines, toxic oxygen radicals, proteases, vasoactive substances) that will ultimately damage the underlying GBM. The result is excessive synthesis of GBM, increased glomerular permeability, development of nonselective proteinuria, and effacement of the foot processes. The reduction in the glomerular filtration rate (GFR) results from progressive thickening of the GBM and reduced glomerular hydraulic permeability coefficient (Kf) whereas the effects of proteinuria on tubular cell differentiation and cytokine release engender progressive tubulointerstitial fibrosis. Megalin, however, is not present in human glomeruli, and except for rare cases secondary to hepatitis B or thyroiditis, where the specific antigen has been found as part of the immune complex, and a well-documented case of congenital MN due to passive transfer of a maternal antibody against a neutral endopeptidase present on the infant's glomerular podocyte, the nature of the antigen involved in the majority of cases of idiopathic MN remains unknown.

It is quite likely that a number of different antigen-antibody combinations may cause MN. Some antigens may be endogenous, while others may be exogenous. The electrical charge of the antigens, antibodies, and/or circulating immune complexes, as well as their size, may play a role in determining whether an immune complex will form in a subepithelial position. There is an increased risk for MN in white patients with HLA-DR3 and in Japanese patients with HLA-DR2. The rare occurrence of MN in more than one family member suggests that genetic factors may be involved in the pathogenesis related either to a predisposition for developing the disease or its progression.

The diagnosis of MN is made by renal biopsy. In very early cases of MN, the glomeruli appear normal by light microscopy (hematoxylin and eosin), although abnormalities may be seen in silver preparations and by immunofluorescence and electron microscopy. Capillary loops are widely patent and the glomeruli show no increase in cellularity and there is no nuclear crowding. As the number and size of subepithelial immune complexes increase, the GBM develops a diffuse and uniform thickening on light microscopy. These changes affect all the glomeruli. Thin sections examined by the periodic-silver-methenamine stain demonstrate the classical “spike” pattern on the epithelial side of the basement membrane reflecting the increased synthesis and deposition of GBM-like material around the immune deposits (Figure 26–1). As the disease progresses, thickening of the capillary wall becomes pronounced, the capillary lumen narrows, and eventually sclerosis and hyalinization of the glomerular tuft develop.

Proximal tubules are remarkable for the lipid vacuoles in the cytoplasm and numerous proteinaceous casts in the lumen. In the initial stages the interstitium is often normal, but with progression of the disease fibrosis and lymphocyte infiltrates are present. Immunofluorescence microscopy shows a very characteristic and uniform deposition of IgG and C3 in a granular pattern along the epithelial side of the GBM (Figure 26–2). IgG (with IgG4 being the predominant IgG subclass) is present in >95% of cases, but C3 may be seen in only 30–50% of cases of idiopathic MN. It has been suggested that positive C3 staining represents active, ongoing immune deposit formation and complement activation at the time biopsy was performed, whereas the absence of C3 reflects cessation of the immunopathologic process. Electron microscopy demonstrates that the typical electron-dense deposits are localized in the subepithelial space together with effacement of the foot processes (Figure 26–3). Deposits usually have a synchronous, homogeneous electron-dense appearance, but a heterogeneous type with dense deposits at various stages of formation can also be found. A four-stage (I–IV; Table 26–2) classification system has been developed based on their specific localization. Unfortunately the clinical and laboratory correlation with these stages is poor.

Secondary forms on MN have histologic features similar to idiopathic MN. However, the presence of deposits of immunoglobulins other then IgG (ie, IgA and IgM), particularly in the mesangium, small subendothelial deposits, tubular basement membrane deposits, and intense C1q deposition are more suggestive of membranous nephropathy secondary to SLE, hepatitis B, or drugs (gold, d-penicillamine).

The disease affects patients of all ages and races, but is more common in men than women by about a 3:1 ratio. Idiopathic MN is most often diagnosed in middle age, with the peak incidence during the fourth and fifth decades of life, and is relatively uncommon in patients under 20 years. At presentation 60–70% of patients will have the nephrotic syndrome (NS) and its associated features: edema, hypoalbuminemia, and hyperlipidemia. The remaining 30–40% of patients present with subnephrotic range proteinuria (<3.5 g/24 hours), most commonly found at the time of routine examination in an otherwise asymptomatic patient. Proteinuria is nonselective. The presence of microscopic hematuria is common (30–40%), but macroscopic hematuria and red cell casts are rare and suggest a different histopathology. In patients with idiopathic MN serum C3 and C4 complement levels are always normal. At the time of diagnosis, the majority of patients are normotensive; only 10–20% are hypertensive. Renal function is normal in the majority of patients at presentation, with only a small fraction (<10%) exhibiting renal insufficiency (Table 26–3) Additional complications related to the disease include a variety of abnormalities in the lipid profile, which probably contribute to the increased cardiovascular risk seen in these patients, and a high prevalence of thromboembolic events including renal vein thrombosis in 10–40% of patients.

The differential diagnosis includes other causes of NS such as minimal change disease, focal segmental glomerulosclerosis, membranoproliferative glomerulonephritis type I and II, amyloidosis, light chain deposition disease, and diabetic nephropathy. It is important to exclude secondary causes of MN, particularly, hepatitis B, SLE, malignancy, and drugs. In patients less than 16 years of age secondary MN is most commonly due to viral infection or SLE, while in adults >60 years, secondary MN is usually due to malignancy or drugs (Table 26–4). In patients with MN, ruling out secondary causes, apart from a thorough history and physical examination, should involve appropriate laboratory evaluation including complement profile, hepatitis serology, antinuclear antibodies, a chest x-ray, testing for occult blood in the stools, a mammogram in women, and prostate-specific antigen testing in men.

In a biopsy series from patients with SLE, MN histology accounts for 8–27% of cases of lupus nephritis. Patients with “pure” membranous lupus nephropathy often have no clinical symptoms that could suggest SLE and serologic markers of lupus activity, such as serum complement and anti-dsDNA levels, are frequently normal and do not correlate with disease activity. Hepatitis B-associated MN occurs frequently in hepatitis B- prevalent areas of the world and affects both adults and children who are chronic carriers of hepatitis B virus [positive hepatitis B surface antigen (HbsAg), hepatitis B core antigen (HbcAg), and usually hepatitis B early antigen (HBeAg)], with or without a history of overt liver disease. In children with hepatitis B-associated MN the nephrotic syndrome usually has a benign course, but progressive renal insufficiency is common in adults.

Hypocomplementemia is present in approximately 50% of the reported cases. Solid tumors (eg, carcinoma of the lung, colon, or kidney) are the most common underlying malignancy involved in cases of tumor-induced MN. It is hypothesized that antigen(s) derived from the tumor are deposited in the glomeruli where they trigger an antibody response and activation of the complement cascade, leading to disruption of the GBM integrity and podocyte injury. In some patients proteinuria resolves with removal or adequate treatment of the tumor. There are, however, well-described cases in which no improvement or remission of the proteinuria occurred following removal of the tumor.

In cases of MN secondary to drugs, discontinuation of the offending agent usually results in complete remission of the nephrotic syndrome. Although resolution of the proteinuria may occur as early as 1 week following discontinuation of the offending drug (eg, NSAIDs), it is not unusual for proteinuria to persist and in certain types associated with the use of gold or d-penicillamine it may take up to 3 years for remission of the proteinuria to occur (mean: 9–12 months). A number of glomerular pathologies have been reported to occur in association with or superimposed upon MN. Such diseases include IgA nephropathy, focal and segmental glomerulosclerosis, crescentic glomerulonephritis (anti-GBM disease, ANCA vasculitis), acute interstitial nephritis, and diabetic nephropathy. In some diabetic patients, MN is thought to occur as a consequence of the development of antiporcine insulin antibodies against porcine insulin deposited along the GBM.

The clinical course is characterized by great variability in the rate of disease progression. Even the natural course is difficult to assess, in part due to the criteria used by the local nephrologists to select patients for biopsy. Historically, spontaneous remissions occur in up to 30% of cases, but this percentage is much lower when patients are selected with higher grades of proteinuria at presentation, eg, proteinuria >8 g/24 hours. In untreated patients the reported 10-year kidney survival has varied from 50% to 70%, although many of these studies have included patients with proteinuria <3.5 g/24 hours producing a bias for a more favorable prognosis. For example, a 72% renal survival at 8 years for 100 untreated patients with MN has been reported. However, 37% of patients were nonnephrotic (proteinuria <3.5 g/day) and in 56% of patients proteinuria was <5 g/day. Furthermore, the median follow-up was 39 months and deaths were excluded from the analysis. Even when these exceptions and the “benign” characteristics of the patients are taken into consideration, 25% reached end-stage renal disease (ESRD) at the end of 8 years. In the majority of the more severely affected patients, the disease progresses, albeit slowly, with up to 40–60% of patients eventually developing ESRD over a 15–20-year span. However, because of its high incidence rate, MN remains the second or third leading cause of ESRD among the primary types glomerulonephritis. Cardiovascular and thromboembolic events are increased in this population, especially in patients who remain nephrotic. When loss of renal function occurs more quickly than expected or there is an unexpected acceleration of the disease, a superimposed condition (eg, interstitial nephritis, anti-GBM disease, renal vein thrombosis) should be considered.

Predictors of Poor Outcome

An accurate predictor of outcome of patients with idiopathic MN would allow more specific targeting of immunosuppressive treatment to those who are at high risk of developing ESRD. However, finding useful markers that predict this group has been difficult. Both age and gender influence outcome, with male sex and increasing age associated with a higher risk for renal failure. As well, the degrees of glomerulosclerosis, tubulointerstitial fibrosis, and vascular disease seen on renal biopsy have all been associated with a poor prognosis. More recently, the percent of glomeruli with focal segmental sclerosis and the configuration of the immune deposits (synchronic/single stage or heterogeneous/multistage) on electron microscopy have been suggested to be predictors of both outcome and treatment response. Neither of these last two indicators has been validated in prospective studies. The degree of renal impairment (creatinine clearance) at presentation is also correlated with long-term renal survival. However, renal function at presentation is widely variable and may be independent of disease severity. A better and more sensitive predictor of long-term prognosis is the ongoing rate of renal function loss as measured by the decline of creatinine clearance over time.

Thus far, the best model for the identification of patients at risk was developed with data derived from the Toronto Glomerulonephritis Registry. This model takes into consideration the initial creatinine clearance, the slope of the creatinine clearance, and the lowest level of proteinuria during a 6-month observation period. This risk score assessment has good performance characteristics and to date is the only one validated in two geographically diverse MN populations, one from Italy and the other from Finland. In the validity data sets the sensitivity of the model varied from 60% to 89%, the specificity from 86% to 92%, and the overall accuracy from 79% to 87%. Based on data using this model, patients who present with a normal creatinine clearance, proteinuria ≤4 g/24 hours, and stable renal function over a 6-month observation period have an excellent long-term prognosis, and conservative treatment only is recommended. Since even in this group the natural history may rarely include worsening of renal function, these patients need to continue to be monitored. Patients whose creatinine clearance remains unchanged during 6 months of observation, but continue to have proteinuria >4 g but <8 g/24 hours, have a 55% probability of developing chronic renal insufficiency and patients with persistent proteinuria >8 g/24 hours, independent of the degree of renal dysfunction, have a 66–80% probability of progression to chronic renal failure within 10 years (Table 26–5).

Most recently, certain levels of elevation of the urinary excretion of α1-microglobulin, β2-microglobulin, IgM, and IgG have also been found to be strong predictors of outcome in MN, but these parameters have yet to be validated.

Relapse after Complete or Partial Remission

About 30% of MN patients will relapse subsequent to a complete remission (CR). The majority who do, however, will relapse only to a subnephrotic range of proteinuria and will have stable long-term function. In the two largest studies of patients with MN who achieved CR, only a few developed, over a long observation period, mild renal insufficiency and none progressed to ESRD. We recently reviewed our data on 350 nephrotic patients with MN and found that the 10-year renal survival was 100% in the CR group, 90% in the partial remission (PR) group, and 45% in the no remission group. Thus, both CR and PR appear to be excellent predictors of long-term renal survival.

Nonimmunosuppressive Therapy

Conservative management is directed at control of edema, treatment of high blood pressure and hyperlipidemia, dietary protein intake, and reduction of proteinuria through inhibition of the renin–angiotensin system. Blood pressure control is important to protect against the cardiovascular risk of hypertension, to reduce proteinuria, and to slow the progression of the renal disease. In the Modification of Diet in Renal Disease (MDRD) study, patients with proteinuria >1 g/day had a significantly better outcome if their blood pressure was reduced to 125/75 mm Hg. Thus, in patients with proteinuric renal disease, including MN, the current target for blood pressure is ≤125/75 mm Hg.

The selection of an antihypertensive agent has been controversial. Numerous studies have shown that angiotensin-converting enzyme inhibitors (ACEI) and/or angiotensin II receptor blockers (ARBs) are cardioprotective and can reduce proteinuria and slow progression of renal disease in both diabetic and nondiabetic chronic nephropathy patients. These classes of drugs reduce glomerular intracapillary pressure and protein ultrafiltration, improve glomerular barrier size selectivity, and have been shown to be renoprotective in experimental models of renal diseases and in both diabetic and nondiabetic renal disease in humans. A recent meta-analysis of some large renal protection trials with ACEI showed that the degree of protection is related to the degree of reduction of proteinuria; if proteinuria is not lowered, the benefit is substantially attenuated. The most recent data from the RENAAL study shows that the renal protective effect of angiotensin II blockade was nearly fully explained by its antiproteinuric effect. Some studies have attempted to address this issue in patients with MN, but they have been small, with a limited follow-up. In some, the use of ACEI has been associated with a significant improvement in the glomerular filtration barrier in patients with MN, but in others the efficacy of ACEI in reducing proteinuria appears to be modest at best (<30% reduction in proteinuria). In those with a positive outcome the antiproteinuric effect of these agents is almost always early (within 2 months of initiation of therapy). Patients at low risk for progression (proteinuria <4 g/24 hours) should be treated with ACE ±ARB since this may further reduce their proteinuria and offer additional renal protection with little chance of significant adverse effects. It is worth noting patients need to be instructed to follow a low salt diet since a high salt intake (eg, 200 mm NaCl/day or 4.6 g sodium/day) can significantly impair the beneficial effects of angiotensin II blockade. Therefore, although these drugs should be tried first, achieving CR in patients with proteinuria >5 g/24 hours using conservative treatment with ACEI and/or ARBs alone appears unlikely, even when these are used at their highest recommended dosages.

Dietary protein intake should be restricted to 0.8 g/kg ideal body weight per day of high-quality protein. Although dietary protein restriction may reduce proteinuria (15–25%) and slow the progression of renal disease somewhat, it has never been shown to induce complete remission of the NS. Proteinuria is also an independent risk factor for cardiovascular (CV) morbidity and mortality. Proteinuric patients have elevated cholesterol and triglycerides and markedly elevated CV risk, as illustrated by the almost 6-fold increase in the incidence of myocardial infarction in this population. It is likely that the lipid abnormalities associated with proteinuria are important players in the high CV risk in these patients, and thus provide an important target for treatment.

Statins have been effective in improving the lipid profile and in reducing CV morbidity and mortality in hyperlipidemic and hypertensive patients and in patients with chronic renal failure. These agents are also effective in reducing serum levels of total cholesterol and low-density lipoprotein (LDL)-cholesterol in nephrotic patients and their use is appropriate in patients with hyperlipidemia. Although no study to date has been conducted to demonstrate that reducing cholesterol lowers the risk of CV events in nephrotic patients, the evidence derived from other studies strongly supports this concept. If the proteinuria persists at >3 g/day even these agents may not completely normalize the lipid profile. The adverse risk profile with these agents in the nephrotic syndrome is similar to the normal population with the exception of an increased incidence of rhabdomyolysis when used in conjunction with high-dose cyclosporine. Apart from the proven efficacy of statins in improving the lipid profile, experimental data together with a number of small, controlled trials have suggested that 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors have a synergistic antiproteinuric effect when combined with ACEI. The antiproteinuric effect of statins is small, and is mainly observed in patients with nonnephrotic range proteinuria. These results are in stark contrast to the reports indicating that proteinuria (albuminuria) may be a complication of statin treatment. These findings have been corroborated by recent phase III studies showing the development of proteinuria (albuminuria) in some patients with the use of high-dose rosuvastatin.

Renal proximal tubule cells are responsible for the reabsorption of proteins present in the tubular lumen. Receptor-mediated endocytosis is the process that is responsible for albumin uptake in proximal tubular cells, a process that requires prenylated GTP-binding proteins. HMG-CoA reductase inhibitors reduce intracellular levels of isoprenoid pyrophosphates that are required for the prenylation and normal function of GTP-binding proteins. Recent studies have demonstrated that statins, as a class, inhibit protein uptake by tubular cells by reducing prenylation of proteins involved in endocytosis and thus inhibit receptor-mediated endocytosis in both animal and human kidney proximal tubular cell lines. These data may help to explain the occurrence of proteinuria in some patients treated with high statin doses. Patients with severe NS are at increased risk for thromboembolic complications. Prophylactic anticoagulation has been shown to be beneficial in reducing fatal thromboembolic episodes in nephrotic patients with MN without a concomitant increase in the risk of bleeding. Although no consensus has emerged as to whether prophylactic anticoagulation should be used in this disease, the majority of physicians would consider using it for patients at high risk of a thromboembolic event. No laboratory test can help predict which patient is in this high-risk category, although thromboembolic events are more common in patients who are severely nephrotic (proteinuria >10g/day and serum albumin <2.5 g/day). Both heparin and low-molecular-weight heparin reduce proteinuria, but this effect has not been routinely employed in the care of patients with MN.

NSAIDs are antiproteinuric but can cause acute renal failure and should be avoided in patients with kidney disease. However, in severe untreatable nephrotic syndrome, NSAIDs can reduce proteinuria by 30–50%, are additive to the effects of ACEI, and can provide symptomatic relief. The combined use of NSAIDs and ACEI/ARBs should be carefully monitored, especially in elderly patients and in those with hypertension and renal insufficiency. In rats with 5/6 nephrectomy, pentoxifylline, a tumor necrosis factor (TNF)-α-suppressing agent, can prevent progression of proteinuria and renal diseases, possibly by suppression of mitogenic and profibrotic genes. In patients with MN, there is increased production of TNF-α in the glomeruli, and urinary excretion of this cytokine correlates with proteinuria. In a pilot study, 10 patients with MN were treated with pentoxifylline (1200 mg/day) for 6 months. Proteinuria decreased from 11 g/day (range 4.6–27) to 1.8 g/day (0–10.9) (p = 0.001), whereas serum albumin concentration increased from 17 g/L to 39 g/L. Complete remission was achieved in five patients. The results are encouraging but are very preliminary.

Lipid peroxidation has been involved in the pathophysiology of proteinuria in Heymann's nephritis model and was the rationale behind using probucol, a lipid peroxidation scavenger, in patients with MN. Treatment with probucol, 1 g/day orally for 3 months, had a modest antiproteinuric effect [proteinuria 6.4 (3.8–9.1) versus 4.7 (1.3–16) g/day pretreatment versus posttreatment (median, range), respectively]. In Heymann's nephritis model, intraperitoneal injection of immune globulin reduced proteinuria by over 50%, accompanied by a reduction in glomerular C5-9 and C3 staining and urinary C5-9 excretion. In humans, intravenous high-dose immune globulin therapy has been beneficial in the treatment of a number of antibody-mediated diseases including idiopathic thrombocytopenic purpura, IgA nephropathy, and ANCA vasculitis. A multitude of mechanisms are thought to be involved in mediating the therapeutic effects of intravenous immune globulin in immune complex-mediated diseases and involve, among others, the blocking of Fc receptors, antiidiotype inhibition of pathogenic immunoglobulin, and the regulation of the complement cascade by preventing the binding of activated C3 to antibody-coated targets, eg, the glomeruli, thus reducing complement-mediated glomerular injury. In patients with idiopathic MN, a short-term low-dose (100–150 mg/kg/day for 6 consecutive days; total dose 600–900 mg/kg) course of intravenous immune globulin was capable of inducing early remission of the nephrotic syndrome (within 6 months), although the benefit was seen mainly in a subgroup of patients with homogeneous type synchronous electrodense deposits on electron microscopy, and long-term renal outcome was unaffected.

Immunosuppressive Therapy

Several treatment strategies, including a variety of immunosuppressive agents, have been shown to be at least partially successful in reducing proteinuria in MN. Many questions about this type of therapy do remain as follows: (1) How long should conservative therapy be extended, and how long should you wait for a spontaneous remission before initiating immunosuppressive therapy, (2) which patients with MN should receive this type of treatment, (3) which of the various drugs available should be used, ie, which are the most effective and safest, and (4) how long should the drug be used before considering it a failure. We will briefly review the available evidence.

Corticosteroids

The early U.S. collaborative study of adult idiopathic nephrotic syndrome reported that a 2–3-month course of high-dose alternate-day prednisone (100–150 mg), when compared to placebo, resulted in a significant reduction in the progression to renal failure, although there was no effect on the degree of proteinuria. The short follow-up period of patients and the worse than expected outcome of the control group have raised criticism of the study. Subsequent controlled studies have shown no benefit from the use of corticosteroids on MN. The two largest randomized controlled trials, one by the British Medical Research Trial and the second by the Toronto Glomerulonephritis Study Group, found no significant benefit of corticosteroid treatment alone in either induction of remission or preservation of renal function. Therefore, the evidence to date does not support the widespread use of oral corticosteroids as a single agent for the treatment of MN.

Cytotoxic Agents Combined with Corticosteroids

In patients with a moderate risk of progression, a significant benefit has been described when a cytotoxic agent alternating monthly with corticosteroids has been used. A number of randomized trials suggest that 6 months of this regimen (cyclophosphamide or chlorambucil as the cytotoxic agent) is four to five times more likely to induce a CR of the NS, and halt disease progression, compared to no therapy or corticosteroids alone. The largest studies with the longest observation time come from Ponticelli's group in Italy. The first study compared the effects of combined methylprednisolone (MTP) 1 g intravenously on the first 3 days of month 1, 3, and 5 followed by 27 days of oral methylprednisolone (0.4 mg/kg/day) or prednisone (0.5 mg/kg/day) alternating in months 2, 4, and 6 with chlorambucil at 0.2 mg/kg/day, versus conservative treatment in 62 patients with MN and nephrotic range proteinuria. The 32 patients who received the alkylating agent were followed for a mean of 31.4 ± 18.2 months. CR was achieved in 50% and PR in 31% of the cases. Among the controls, CR was achieved in 7% and PR in 23% of the patients. The regimen was remarkably safe with only four of the treated patients stopping therapy. After up to 10 years of follow-up, patients treated with combination therapy had a 92% probability of renal survival compared with 60% in the control group and only 8% of treated patients versus 40% of untreated ones had reached ESRD. In addition, the slope of the reciprocal of serum creatinine remained significantly slower in the treated group than in the untreated controls for up to 90 months. In terms of proteinuria, only 42% of the treated group, versus 78% of the placebo group, spent time in a nephrotic state over the 10-year follow-up. Women and patients with mild glomerular lesions (stage I and II) were more likely to enter remission after combined therapy in this study, and no patients had significantly impaired renal function at entry.

A second study compared 6 months of alternating monthly pulses of MTP (1 g), oral steroids, and chlorambucil as described above versus MTP pulses and steroids (0.4 mg/kg every other day) alone found that at 3 years, 66% of the patients given steroids and chlorambucil versus 42% of the patients given steroids alone were in remission, the difference being significant. At 4 years, however, this difference was no longer statistically significant, although a seemingly large 20% difference favoring the combined treatment still persisted. Combined therapy was also associated with a trend toward better preservation of renal function, as assessed by serum creatinine, although again the difference was not statistically significant.

In a third study from the same group, patients were enrolled in a 6-month study comparing MTP/prednisone alternating with chlorambucil (same doses as the prior studies) to MTP alternating with oral cyclophosphamide (2.5 mg/kg/day). Among 87 nephrotic patients followed for at least 1 year, 82% of patients assigned to MTP and chlorambucil had complete or partial remission of the NS versus 93% of patients assigned to MTP and cyclophosphamide (p = 0.116, NS). The use of cyclophosphamide was associated with fewer side effects, but renal function was equally preserved in both groups for up to 3 years. However, a relapse rate in both treated groups of 25–30% was seen within 2 years. The incidence of CR 1 year following combined immunosuppressive treatment was approximately 28% in the first study, 20% in the second study, and 32% in the third study. In addition, these studies excluded all patients whose serum creatinine exceeded 1.7 mg/dL at entry.

More recently, in an uncontrolled trial, 39 patients with severe NS (proteinuria 8.9 ± 3.6 g/24 hours) and increased serum creatinine at entry (>1.5 mg/dL; creatinine clearance <60 mL/minute) were treated with chlorambucil (0.15 mg/kg/day) for 14 weeks plus oral prednisone 1 mg/kg tapered to 0.5 mg/kg/every other day over a 6-month period and were compared to a similarly affected historical control group from the 1970s. After 4 years of follow-up, there was a 90% probability of renal survival without dialysis in those who received treatment compared with only a 55% probability in those receiving conservative treatment. After 7 years, the renal survival was 90% and 20%, respectively.

These observations have been recently confirmed by Jha et al. who reported the 10-year follow-up of a randomized, controlled trial on 93 patients allocated to either conservative therapy or to receive a 6-month course of alternating prednisolone and cyclophosphamide. Proteinuria was 5.9 ± 2.2 and 6.1 ± 2.5 g/24 hour in the conservative and immunosuppressive therapy group, respectively. Renal function was well preserved with estimated GFR rates above 80 mL/minute in both groups. Of the 47 patients treated with immunosuppressive therapy, 34 achieved remission (15 C and 19 PR), compared with 16 (5 C, 11 PR) of 46 in the control group (p < 0.0001). The 10-year dialysis-free survival was 89 and 65% (p = 0.016), and the likelihood of survival without death, dialysis, and doubling of serum creatinine was 79% in the treated versus 44% (p = 0.0006) in the control group. The incidence of infections was similar in the two groups.

A recent meta-analysis of the available literature showed that the use of alkylating agents was associated with higher rates of remission (PR or CR), but there was no statistical difference when compared to placebo in ESRD or the death rate. Results of a meta-analysis need to be viewed with caution, since this technique is not a substitute for a well-designed randomized control trial of adequate size.

A number of uncontrolled studies have examined the effects of chlorambucil or cyclophosphamide on the outcome of MN in patients with advanced disease (serum creatinine >1.8 mg/dL). In one study eight patients with deteriorating renal function were treated with a modified Ponticelli's protocol using a lower chlorambucil dose (0.15 mg/kg/day) alternating monthly with cycles of prednisone. Proteinuria was reduced in all patients; creatinine clearance increased in six patients and the rate of renal function decline was reduced in the other two patients. However, adverse effects of chlorambucil were severe and patients with severe renal insufficiency (serum creatinine >3 mg/dL) had progressive deterioration in function.

Similarly, in another study 21 patients with heavy proteinuria and progressive renal failure (creatinine 2.0–5.4 mg/dL) were treated with alternating monthly cycles of prednisolone (125 mg on alternate days given in months 1, 3, and 5) and chlorambucil (10 mg/day in months 2, 4 and 6). In addition, nine patients received at least one course of three intravenous pulses of 1 g MTP at the start of the first cycle of treatment. After a mean period of follow-up of 39 months, three patients had died, six patients were on dialysis or had a serum creatinine ≥5.7 mg/dL, and one patient had progressive renal failure. Eleven patients had either stable or improved renal function, as judged by serum creatinine concentration. Treatment was associated with a high incidence of side effects, with significant complications related to drug therapy observed in >50% of subjects.

In an uncontrolled study, 32 high-risk patients were treated with monthly pulses of MTP (1 g intravenously on 3 consecutive days) followed by oral prednisone (0.5 mg/kg/day in months 1, 3, and 5) and chlorambucil (0.15 mg/kg/day in months 2, 4, and 6) (n = 15) or oral cyclophosphamide (1.5–2.0 mg/kg/day for 12 months) and steroids in a comparable dose (n = 17). All patients had evidence of deterioration of renal function prior to the start of therapy; serum creatinine was 2.5 ± 0.8 in the chlorambucil group and 3.0 ± 1.4 mg/dL in the cyclophosphamide group. Treatment resulted in rapid improvement in renal function in both groups (serum creatinine 1.9 ± 0.7 mg/dL in both groups at 6 months) but the improvement was short-lived in the chlorambucil group. At 6 months, there was a significant decrease in proteinuria in both groups, but again, the decrease was persistent only in the cyclophosphamide group (chlorambucil group 9.1 ± 2.6 g/day versus 6.8 ± 4.4 g/day and cyclophosphamide group 11.2 ± 5.3 g/day versus 2.0 ± 3.0 g/day at months 0 and 12, respectively). Overall, a partial remission of proteinuria was observed in five (33%) patients after chlorambucil treatment and in 15 (92%) patients after cyclophosphamide treatment (p < 0.01). Of these latter patients, six had complete remission, whereas none of the chlorambucil-treated patients did. Adverse effects were common and treatment had to be reduced, temporarily interrupted, or prematurely stopped in 6 of 17 cyclophosphamide-treated patients and in 11 of 15 chlorambucil-treated patients.

In contrast, in the only randomized study of patients at high risk of progression (mean creatinine 2.3–2.7 mg/day; proteinuria 11.1–12.5 g/day), 13 patients were assigned to monthly pulse cyclophosphamide combined with oral prednisone for 6 months and were compared to 13 patients treated with prednisone alone for the same period of time. At the end of follow-up, there were no statistical differences regarding degree of proteinuria, frequency of remissions, or rate of decline of renal function between the steroid-alone and the combined-treatment group. Other small uncontrolled studies using cyclophosphamide alone have found similar results. These results clearly differ from those discussed above. One possible explanation is that daily cyclophosphamide is more effective than cyclophosphamide used in monthly intravenous pulses.

To address the long-term efficacy of cytotoxic therapy in these patients, 65 patients with MN and serum creatinine >1.5 mg/dL who were treated with oral cyclophosphamide (1.5–2.0 mg/kg/day for 12 months) and steroids (MTP pulses of 3 × 1 g intravenously at months 1, 3, and 5, and oral prednisone 0.5 mg/kg on alternate days for 6 months) were prospectively studied. Follow-up was 51 (5–132) months. Renal function improved or stabilized in all patients. Overall renal survival was 86% after 5 years and 74% after 7 years. A PR occurred in 56 patients followed by a CR in 17 patients. However, 11 patients had a relapse (28% relapse rate at 5 years), of whom 9 were retreated because of deteriorating renal function. Treatment-related complications occurred in two-thirds of patients, mainly consisting of bone marrow suppression and infections.

Thus, in a number of studies, both cyclophosphamide and chlorambucil in combination with corticosteroids appear to be effective in the treatment of patients with idiopathic MN and preserved renal function. This combination may also be effective in those with deteriorating renal function, but the supporting data are much less compelling, adverse effects are higher, and the likelihood of benefit is reduced in patients with severe renal failure (serum creatinine >3 mg/dL). The favorable effects are maintained well beyond the 1-year treatment period but relapse rates approached 35% at 2 years.

The adverse effects of cyclophosphamide when used long term are the major drawbacks to the universal application of this form of therapy. These include increased susceptibility to infections, anemia, thrombocytopenia, nausea, vomiting, sterility, and in the long-term secondary malignancy, in particular bladder cancer. In regard to chlorambucil, the major concern is the possibility of inducing acute leukemia or lymphoma. Even when chlorambucil was used in modified versions (lower doses) of the Italian regimen, the rate of adverse effects was high.

Cyclosporine

Early uncontrolled studies of cyclosporine suggested an initial benefit but a high relapse rate. In a recent single blind randomized controlled study, 51 patients with steroid-resistant MN were treated with low-dose prednisone plus cyclosporine A (CsA) compared to placebo plus prednisone. CsA was given at 3.5 mg/kg/day with a target whole-blood trough level of 125–225 ng/mL. All patients received prednisone at 0.15 mg/kg/day (up to a maximum of 15 mg/day). At the end of treatment at 26 weeks, 75% (21 of 28 patients) in the CsA group versus only 22% (5 of 23 patients) of controls had achieved a PR or CR (CR = 2 in the CsA group versus 1 in the placebo group). CsA was well tolerated, and no one had to discontinue treatment because of adverse effects. Relapses occurred in ˜40% of patients within 1 year of discontinuation of CsA treatment, not dissimilar to what was found in controlled cytotoxic/corticosteroid regimens. There has been only one randomized controlled trial using CsA in patients with high-grade proteinuria and progressive renal failure, again conducted by the Toronto group. In this study, 64 patients were placed on a restricted protein diet (<0.9 g/kg) and followed for 12 months (phase 1). Seventeen patients with a loss in creatinine clearance of ≥8 mL/minute/year (but with creatinine clearance ≥30 mL/minute) and proteinuria ≥3.5 g/24 hours were randomly assigned to either CsA treatment (3.5 mg/kg/day; nine patients) or placebo (eight patients) for 12 months (phase 2). After 12 months, there was a significant reduction in proteinuria and in the rate of loss of renal function in the CsA group compared with the group that received placebo. In the CsA group, the slope of creatinine clearance was reduced from −2.4 to −0.7 mL/minute/month, whereas in the placebo group the change was insignificant, −2.2 to −2.1 mL/minute/month (p < 0.02). This improvement was sustained in ˜50% of the patients for up to 2 years after CsA was stopped.

That prolonging the treatment results in a higher and more sustained rate of remission is supported by data from the German Cyclosporine in Nephrotic Syndrome Study Group. In this study, 41 high-risk patients with MN and proteinuria >3.5 g/24 hours (mean 10.9 ± 5.7 g/24 hours) were treated with CsA (average dose = 3.3 ± 1.1 mg/kg/day) for a median of 353 days (159–586 days). Approximately 65% of the patients also received ACEI plus corticosteroid treatment (˜27.5 ± 21.2 mg/day). CR (proteinuria <0.5 g/24 hours) was achieved in 34% of the patients. The median treatment time to CR was 225 days (quartiles 120 days and 459 days). Taken together, these data would suggest that CsA will induce a remission (CR or PR) of the NS in 50–60% of patients. It is important to emphasize that although reduction of proteinuria usually occurs within a few weeks, the majority of CR occurred after more then 6 months of treatment. This would suggest that if urinary protein excretion is not significantly reduced within 3–4 months of initiating therapy it is unlikely that more prolonged therapy will result in a remission. This is an important observation and may explain why studies in which CsA was used for less then 6 months achieved low rates of CR whereas studies using CsA for up to 1 year, albeit uncontrolled, reported remission rates close to 80%. However, significant adverse effects including hypertension, gingival hyperplasia, gastrointestinal complaints, muscle cramps, and most important nephrotoxicity can accompany prolonged CsA treatment. The latter is dose/duration dependent as well as age dependent. Patients at particular risk are those with initial impaired renal function, especially if accompanied by chronic vascular plus or minus tubulointerstitial damage on renal biopsy. On the other hand, prolonged low-dose CsA (˜1.5 mg/kg/day) could be considered for long-term maintenance of patients with preserved renal function who achieve a CR or PR, but then relapse once CsA is discontinued, with little risk of nephrotoxicity.

Tacrolimus

Tacrolimus is a calcineurin inhibitor similar to CsA, but it is 50–100 times more potent as an immunosuppressive agent, on a molar basis, than CsA at suppressing lymphocyte proliferation in vitro and graft rejection in vivo. In animal models, pretreatment with tacrolimus prevented the development of experimental MN and significantly reduced urinary protein excretion in animals with established MN. The relevance of these data to human membranous nephropathy is unknown and data on the use of tacrolimus to treat human glomerulonephritis are limited. We have treated 10 patients with MN and proteinuria >5 g/24 hours with at least a 6-month course of tacrolimus monotherapy. Proteinuria decreased to subnephrotic levels in the majority of patients, but no patient achieved CR of the NS (unpublished observations).

A recent study by Praga et al. evaluated tacrolimus monotherapy in MN. In this study 25 patients with normal renal function (mean proteinuria ˜8g/24h), received tacrolimus (TAC; 0.05 mg/kg/day) over 12 months with a 6-month taper, whereas 23 patients served as control. After 18 months, the probability of remission was 94% in the tacrolimus group but only 35% in the control group. Six patients in the control group and only one in the tacrolimus group reached the secondary end point of a 50% increase in their serum creatinine. Unfortunately, almost half of the patients relapsed after tacrolimus was withdrawn, and similar to patients treated with CsA, maintenance of remission may require prolonged use of tacrolimus in low doses.

Mycophenolate Mofetil

Mycophenolate mofetil (MMF) has been used as an effective antirejection agent in solid organ transplants for >10 years. It is metabolized to mycophenolic acid, the active immunosuppressant compound, and selectively inhibits T- and B-lymphocyte proliferation through inhibition of de novo purine synthesis and inactivation of inosine monophosphate dehydrogenase. Mycophenolate mofetil directly suppresses the synthesis of antibodies by B cells and the generation of cytotoxic T cells, and decreases the expression of adhesion molecules on lymphocytes, thus impairing the ability of activated lymphocytes to bind to endothelial cells. The latter effect may reduce the influx of inflammatory cells into glomeruli after deposition of antibody. It does not share azathioprine's profile in regard to myelotoxicity, hepatotoxicity, and mutagenesis. Unlike cyclosporin A or tacrolimus, MMF is not nephrotoxic. Recognized side effects include nausea, vomiting, diarrhea, abdominal pain, anemia, and leukopenia. As with any immunosuppressive agent, patients on MMF are at risk of infections.

There has been a paucity of studies using MMF in MN. In a pilot study, 16 patients were treated with MMF, 1.5–2 g/day, for a mean of 8 months. These patients would be categorized as either medium or high risk for progression given the severity of their proteinuria and the fact that they had previously not responded to a variety of other immunosuppressive drugs. The results were modest: Six patients had a ≥50% reduction in their proteinuria, two had a minor reduction in proteinuria, four had no change, three were withdrawn because of significant adverse effects, and one stopped treatment on his own. There were no significant changes in mean serum creatinine, or serum albumin levels, over the course of the study. In patients who responded, the lowest degree of proteinuria was reached within 6 months, suggesting that patients who are likely to respond would do so in this time frame. This was a pilot study, and is somewhat difficult to interpret as negative or positive, given the setting of resistance to all other agents.

Similar results were reported in a retrospective analysis of 17 patients with MN (15 patients had nephrotic range proteinuria and 6 had renal insufficiency). Patients were either steroid dependent, steroid resistant, or steroid intolerant, with or without CsA, or had been resistant or had a suboptimal response to CsA, or had signs of progressive renal failure. Overall, there was a 61% reduction of proteinuria (7.8–2.3 g/24 hours; p = 0.001), with eight patients having a PR and two patients a CR. This group of MN patients was much more heterogeneous in terms of either a prior response to therapy or drug dependency before the initiation of MMF treatment.

More recently, Branten et al. reported on 32 patients with MN and renal insufficiency (serum creatinine >1.5 mg/dL) treated with MMF (1g twice daily) for 12 months and compared with results obtained on 32 patients from a historic control group treated for the same period of time with oral CYC (1.5 mg/kg/day). Both groups received high dose steroid treatment (MTP IV 1g × 3 on months 1, 3, and 5 followed by oral prednisone 0.5 mg/kg every other day for 6 months, with subsequent tapering). Overall, 21 MMF-treated patients developed PR of proteinuria; in 6 patients, proteinuria decreased by at least 50%, and no response was observed in five patients. Cumulative incidences of remission of proteinuria at 12 months were 66% in the MMF group versus 72% in the CYC group (p = 0.3). Side effects occurred at a similar rate between the two groups but relapses were much more common in the MMF treated group.

Rituximab

There is convincing evidence from both experimental and human studies that MN is mediated by the deposition of IgG antibodies in the subepithelial aspect of the GBM. Given the key role of IgG antibodies in MN, it is reasonable to postulate that suppression of antibody production by depleting B cells and/or plasma cells may improve or even resolve the glomerular pathology. Rituximab is a genetically engineered, chimeric, murine/human IgG1κ monoclonal antibody against the CD20 antigen found on the surface of normal and malignant pre-B and mature B cells, but not expressed on hematopoietic stem cells, pro-B cells, normal plasma cells, or other normal tissues. The Food and Drug Administration approved it in 1997 for the treatment of relapsed or refractory low-grade or follicular, CD20+, B cell non-Hodgkin's lymphoma. In a pilot study using rituximab in idiopathic MN, eight nephrotic patients with MN were prospectively treated with a 4 weekly course of Rituxan (375 mg/m2) and followed for 1 year. All patients had complete depletion of circulating B cells lasting up to 1 year. Proteinuria significantly decreased from a mean ± SD of 8.6 ± 4.2 g at baseline to 4.3 ± 3.3 g (−51%, p < 0.005) at 3 months, 4.0 ± 3.1 g (−53%, p <0.005) at 6 months, and to 3.0 ± 2.5 g (−66%, p < 0.005) at 12 months. At 12 months, proteinuria decreased to <0.5 g/24 hours in two patients and <3.5 g/24 hours in three other patients. Proteinuria decreased in the remaining patients by 74%, 44%, and 41%, respectively. Renal function remained stable in all patients. Adverse effects were reported as mild and included chills and fever in one patient and an anaphylactic reaction in another patient. This pilot study, although encouraging, needs to be confirmed by other studies before recommendations can be made regarding its use.

We recently conducted a prospective open-label pilot trial in 15 newly biopsied patients (<3 years) with IMN and proteinuria >4 g/24 hours despite ACEi/ARB use for >3 months and systolic BP <130 mm Hg. Thirteen males and 2 females, median age 47 (range 33–63), with a mean serum creatinine of 1.4 ± 0.5 mg/dL were treated with rituximab (1 g) on day 1 and 15. At six months, patients who remain with proteinuria >3 g/24 hours and in whom total CD19+ B cell count was >15 cells/μL received a second identical course of rituximab. All patients tolerated rituximab well, and achieved swift B lymphocyte depletion by day 28. Baseline proteinuria of 13.0 ± 5.7 g/24 hours (range 8.4–23.5) decreased to 9.1 ± 7.4 g, 9.3 ± 7.9 g, 7.2 ± 6.2 g and 6.0 ± 7.0 g/24 hours (range 0.2–20) at 3, 6, 9 and 12 months, respectively (mean ± SD). Fourteen patients completed a 12 months follow-up: complete remission (proteinuria <0.3 g/24 hours) was achieved in 2 patients, partial remission (<3 g/24 hours) in 6 patients, and 5 patients did not respond. Two patients progressed to ESRD. The mean drop in proteinuria from baseline to 12 months was 6.2 ± 5.1 g and was statistically significant (p = 0.002, paired t-test). Rituximab was well tolerated, and was effective in reducing proteinuria in patients with IMN. However, the responses varied widely among patients and further research is needed in order to identify a priori which patient is likely to benefit from rituximab treatment. A recent study has also shown that a single dose of rituximab is also effective in inducing remission in some patients. We are currently conducting a new study of 20 patients with MN being treated with rituximab (375 mg/m2) weekly for 4 weeks where we have incorporated a number of mechanistic studies to see if we can predict response in these patients.

Eculizumab

The proinflammatory activities of the activated terminal complement components C5a and C5b-9 have been implicated in a wide range of inflammatory disease states including contributing to the glomerular injury in MN. Finding the C5b-9 membrane attack complex within the immune deposits and recognizing that experimental depletion of complement by cobra venom serum prevents subsequent proteinuria confirm the role of complement in the pathogenesis of MN. Eculizumab is a new, humanized anti-C5 monoclonal antibody designed to prevent the cleavage of C5 into its proinflammatory byproducts. In a recent randomized controlled trial (currently reported in abstract form only), 200 patients with MN were treated every 2 weeks with two different intravenous dose regimens and compared to a placebo group over a total of 16 weeks. Neither of the active drug regimens of eculizumab showed any significant effect on proteinuria or renal function when compared to placebo. It was later determined that adequate inhibition of C5 was seen in only a small percentage of patients, suggesting that the doses given were inadequate. More encouraging results were seen in a continuation of the original study in which eculizumab was used for up to 1 year, with a significant reduction in proteinuria in some patients (including two patients who went into complete remission). Whether complement inhibition with higher doses of eculizumab will prove to be more effective, as well as safe, in the treatment of MN remains a question for the future.

Idiopathic MN is a glomerular disease usually of abrupt onset and is associated with the NS. Control of the NS, specifically a CR or PR, is clearly associated with prolonged renal survival and a slower rate of progression of renal disease. There are no standard or universal first-line specific therapeutic options for idiopathic MN. Supportive or conservative care should be given in all cases and should include the use of diuretics and antihypertensive (and potentially renal protective) agents such as ACEI and ARB therapy and lipid-lowering agents. In patients who are at low risk of progression, defined by normal renal function and proteinuria ≤4 g/day over a 6-month observation period, a conservative approach should suffice, given their excellent prognosis (<5% risk for progression over 5 years of observation). These patients need to be followed long term to ensure that there is no disease progression.

Patients at medium risk of progression, defined by normal renal function and persistent proteinuria between 4 and 8 g/day over a 6-month observation period despite maximum conservative treatment, or at high risk of progression, defined by deteriorating renal function and/or by persistent high-grade proteinuria ≥8 g/day during the 6 months of observation, are candidates for immunosuppressive therapy. It is worth emphasizing that patients with persistent nephrotic range proteinuria exhibit marked abnormalities in their lipid profile and are at increased risk for CV complications. Proteinuria is a marker for predicting CV risk as demonstrated by data from the Framingham study showing that proteinuria predicts CV outcome. Data from recent studies (eg, LIFE and the AASK trials) show that CV risk increases with each increase in the level of proteinuria, making the link between chronic renal disease and cardiac disease so strong that few patients develop chronic kidney disease without clinically apparent or occult cardiac disease. The hypothesis that proteinuria is not only an indicator but is an inducer of kidney injury is supported by a number of studies showing that proteinuria plays a major role in the development of progressive tubular injury, interstitial fibrosis, and subsequent loss in GFR. There is overwhelming clinical evidence that higher sustained levels of proteinuria predict more rapid decline in renal function, more pronounced tubulointerstitial injury, and eventual kidney failure. Nephrotic patients are also at risk for thromboembolic events, with an incidence as high as 50% in patients with severe MN being reported. These events are associated with a mortality rate as high as 42% in high-risk patients.

These data emphasize that these life-defining events, in addition to the potential renal failure, are common in these patients. Therefore, even if the main benefit of immunosuppressive therapy is to speed up the induction of a remission that may have occurred spontaneously, it may still have value in the long term. A treatment algorithm that combines the predictive factors and best evidence for immunosuppressive therapy is presented in Figure 26–4. The recommendation to use cytotoxic/steroid or cyclosporine as the first line of therapy in the moderate or high-risk group is based on evidence from trials conducted in patients in these respective categories, but physicians must take into account individual patients and their wishes in order to decide which therapy should be initiated. These routines are not mutually exclusive and may follow sequentially (with a drug holiday) if the first one chosen does not succeed in reducing the proteinuria to the desired range and/or adverse side effects make completion of a course of therapy untenable. Patients who do not respond well or relapse after a first course of immunosuppression therapy may benefit from a second course of immunosuppression.

Preliminary evidence on the use of anti-CD20 antibodies suggests this is another agent that may be as effective, and safer, than our current regimens, but it needs to be assessed further before being widely recommended. Patients with severe renal insufficiency (serum creatinine ≥3 mg/dL) are less likely to benefit from immunosuppression therapy and the risk of treatment is significantly higher. These patients should be considered for conservative therapy only and plans should be made for transplantation in the future.