Extramembrane glomerulopathy

Extramembrane glomerulopathyPhysiopathology:

Extramembraneous glomerulonephritis (GEM) is characterized by deposits on the outer side of the glomerular basement membrane. The antigenic target is located on the podocyte. Proteinuria appears after the formation of the deposits, which are associated with the flattening of the podocytes and their erasure. Thereafter, the glomerular basement membrane thickens with extracellular matrix accumulation.

Heymann nephritis is the experimental model of GEM in rats. In this model, the identified antigenic targets are megalin and the protein complex associated with megalin (RAP for receptor associated protein). Megalin is a transmembrane protein of more than 600 kD that acts as an endocytosis receptor, binding multiple ligands such as calcium, E, J, and B 100 apoproteins, the urinary plasminogen activator inhibitor complex, antibiotics , lactoferrin, albumin, transcobalamin-vitamin B12 complex, vitamin D binding protein and possibly insulin In the kidney, it is expressed in clathrin-coated wells on the surface of podocytes and in the brush border of the proximal tubular cells. It is also present in pulmonary type II cells. It belongs to the family of low density lipoprotein (LDL) receptors. The RAP complex binds to megalin during the endocytosis process. RAP is essentially present in the endoplasmic reticulum of podocytes. The formation of the complex inhibits the binding of megalin to its ligands. There is a single epitope on RAP, with a length of 14 amino acids.

The injection of an antibody directed against this epitope made it possible to reproduce deposits of immune complexes. However, many other pathological epitopes on megalin are necessary for the in situ formation of immune complexes. The epitopes of megalin are located in the four extracellular domains rich in cysteine. The megalin and RAP combination is referred to as the antigenic complex of Heymann’s nephritis.

Heymann nephritis can be induced by intravenous injection of megalin or RAP or, passively, by injection of antimalin and anti-RAP antibodies. In the passive model, deposits are detected a few minutes after injection and proteinuria appears 5 to 6 days later. The deposits form initially in the clathrin-coated wells at the base of the podocytes, and then extend on the outer side of the glomerular basement membrane. The mechanisms of this extension towards the basal membrane, which is itself devoid of megalin, are not known.

The formation of immune complexes on the membrane surface of podocytes is responsible for complement activation and insertion of the C5b-9 membrane attack complex into the podocytic membrane via the vitronectin receptor. This attack complex plays a central role in podocytic lesions and in the appearance of proteinuria. Its role has also been demonstrated in humans. Treatment with venom of cobra after induction of Heymann nephritis makes it possible to block the appearance of proteinuria by supplemental depletion, even though deposits of immune complexes are present on the podocytic surface. The podocyte responds to the membrane insertion of C5b-9 by endocytosis of this complex, probably a cell defense mechanism. The endocytosis vesicles are released into the urinary space, explaining the presence of C5b-9 in the urine of patients with GEM. In an advanced stage of Heymann’s nephritis, associated with the development of proteinaceous, attack complexes are found in the podocytic membrane, which could be the result of a limitation of the endocytosis system.

In response to the insertion of C5b-9 into the podocyte membrane, many enzymatic activities are altered in the podocytic cell. A cascade of oxidation activation will lead to the appearance of proteinuria. The enzyme oxidoreductase nicotinamide adenosine dinucleotide phosphate (NADPH) is activated and the antioxidant effect of glutathione peroxidase and superoxide dismutase is decreased. Treatment with vitronectin, which blocks the insertion of C5b-9 after induction of Heymann nephritis, blocks the activation of oxidation in the podocytic cell. Oxidation generates free radicals. Free radical scavengers, administered in Heymann nephritis, reduce the intensity of proteinuria without modifying the intensity of the deposits. Free radicals in turn generate lipid oxidation. Membrane lipids are not those that are directly oxidized, as evidenced by the absence of oxidized phosphatidylcholine in the membrane. Large quantities of apolipoprotein E (apoE) and apoB are found bound in the immune complexes to megalin, receptor of these two apolipoproteins. These apolipoproteins are subjected to the increase in the oxidative power of the podocyte. The lipid oxidation products of these apolipoproteins bind to collagen IV on the NC1 domain rich in cationic residues and could be responsible for the proteolytic degradation of the basement membrane.Probucol, which blocks lipid oxidation, has been shown experimentally to reduce proteinuria in Heymann’s nephritis by 85%.

It is not known how to explain the changes in permeability of the glomerular basement membrane as a result of these phenomena. The structural modifications of collagen IV but also the oxidation of apolipoproteins can contribute to this.The precise role of proteinase modifications in the increase in membrane permeability is not known.

The insertion of C5b-9 further induces an increase in the activity of gelatinase, metalloproteinase 9, fibroblast growth factor, phospholipase C, A2 system and protein kinase C. Thickening of the glomerular basement membrane occurs secondary to the accumulation of extracellular matrix proteins. The transforming growth factor b (TGF- b ) could act as a mediator in this accumulation.

Etiologies and associations:

In children, infectious causes and systemic lupus erythematosus are the most frequent etiologies, while the neoplastic and medicinal causes predominate in adults.

Infectious agents such as malaria, bilharziasis and hepatitis B are probably the most common causes in the world. In Asia, hepatitis B is listed as the most common etiologic agent. In Taiwan, where 20% of the population is HBsAg-positive, 96% of children with GEM are HBsAg-positive. In Japan, where the prevalence is only 2-3%, 57% of children with GEM are HBsAg-positive. In the United States, the prevalence of HBsAg is 0.3% in the general population and 20% in children with GEM. In a population in contact with a carrier of the HBs antigen, the frequency of proteinuria increases with respect to a control population.

The proper pathogenic role of the Hbs antigen is discussed, its high molecular weight not allowing it to cross the glomerular basement membrane. It is most often absent from immune deposits in immunofluorescence. The Hbe antigen, with a molecular weight of 30,000 daltons, is more likely the pathogen. It is most frequently found in extramembrane deposits.

In this etiological form, patients may have hypocomplementemia (27 to 91% of patients). The prognosis is relatively favorable, with 8% of end-stage renal disease in a series of adults. The non-remission factors identified are age greater than 6 years, duration of proteinuria greater than 12 months, absence of seroconversion for HBeAg, histological stage II and III. Interferon alpha, previously recommended for the treatment of hepatitis B, had no recognized therapeutic effect on hematitis B (HBV) -based GEM. However, cases of regression of GEM under antiviral treatment have been published.

In GEM associated with human immunodeficiency virus (HIV) infection, a case of remission was published after initiation of antiretroviral therapy. A response to corticosteroid therapy has also been reported in this etiology.

Nephrotic syndrome secondary to nonsteroidal anti-inflammatory drugs (NSAIDs) is most often associated with glomerulopathy with minimal glomerular lesions, but the occurrence of GEM is not uncommon. Radford shows that out of 125 patients with GEM, 29 had NSAIDs and 13 met criteria for drug accountability, including a rapid regression of proteinuria when it stopped. All classes of NSAIDs are involved: diclofenac, fenoprofen, ibuprofen, flurbiprofen, oxaproxin, sulindac, nabumetone, naproxen, piroxicam and tolmetin.

Mercury is a very volatile liquid metal even at room temperature. It may be absorbed by inhalation, ingestion or through the skin. Its toxicity is related to enzymatic inactivation which leads to acute toxicity of interstitial pneumonia or erosive gastroenteritis, through the intoxication pathway, and tubular necrosis. In chronic poisoning, mercury mainly affects the central nervous system and the kidney by an extramembrane involvement. The old exposures reported are mercurial diuretics, toothpaste and creams containing mercury against psoriasis. Mercury is still used today in the industry especially in batteries, fluorescent tubes and in measuring devices. Creams used to whiten the skin contain mercury up to 2,000 times the permitted level and have been incriminated in the occurrence of GEM.

Formaldehyde poisoning occurs in homes where formaldehyde has been used in assembly resins or formerly in insulation. Air usually contains less than 0.01 particle per million (ppm) and may contain up to 2 ppm in homes or mobile homes where formaldehyde has been used. Starting at 0.05 ppm, a strong smell is felt and signs of eye and respiratory irritation may appear.

Exposure to formaldehyde has also been responsible for urticaria, dermatoses and ENT cancers.

WHO Class V of lupus kidney disease refers to GEM. It accounts for approximately 14% of systemic lupus erythematosus nephropathies (LEAD) if cell-proliferative classes are included and less than 10% without proliferative forms. The old classification grouped the GEMs without cell proliferation (pure, Va) and the GEMs associated with diffuse extracapillary proliferation (Vd). In the new classification of the World Health Organization (WHO), subclasses Vc and Vd were deleted and reinstated in classes III and IV, of which they share the prognosis. GEM may precede the occurrence of clinical LEAD and the appearance of antinuclear factors. Certain histopathological signs, including the extension of the immune deposits on the endocapillar and tubular slopes, the nature of the deposits and the existence of mesangial cell proliferation with extracapillary, may make this etiology suspect. The deposits commonly associate IgG, IgM, IgA and C1q. Hypocomplementemia occurs in 6 to 59% of cases with mostly low levels of anti-deoxyribonucleic acid (DNA) antibody, low precipitating capacity of these antibodies and few circulating immune complexes. The literature reports a renal survival of 80 to 90% at 10 years, the pejorative evolution may be related to a fibrosing evolution comparable to the idiopathic form or to a transformation into a proliferative lupus nephropathy type III or IV.

Hypocomplementemic urticarial vasculitis is most often complicated by mesangioproliferative glomerulopathy, but cases of GEM have also been reported.

Hashimoto’s thyroiditis as well as Grave’s disease have been described in association with a GEM. Renal lesions were reproduced in the rabbit by injecting thyroglobulin, with proliferation of endothelial cells and thickening of the glomerular basement membrane. The presence of thyroglobulin has been demonstrated along the glomerular basal membrane in humans during GEM secondary to hyperthyroidism. In a series of 84 patients with autoimmune thyroiditis,

30 had an abnormal protein / creatinine urinary ratio. There was no demonstrated relationship between the level of autoantibodies and the existence or absence of proteinuria in this series. Cases of GEM after treatment with radioactive iodine have also been reported. Weetman records the occurrence of proteinuria in nine out of 14 patients with Graves’ disease treated with radioactive iodine and without proteinuria prior to treatment. It is possible that the release of thyroglobulin during the treatment with radioactive iodine leads to the development of GEM.

Patients with rheumatoid arthritis may develop GEMs induced by gold salts or D-penicillamine, but also spontaneously.Nakano counts nine patients without treatment before the appearance of this type of kidney damage.

GEMs with extracapillary involvement outside LEAD have been described in association with glomerular basal antimembrane antibodies or anticytoplasm of neutrophil polymorphonuclear cells. In children, GEMs are described in combination with a tubular basal antimembrane antibody. This form is characterized by a complete or incomplete Fanconi syndrome and by the possibility of extrarenal damage such as alveolar haemorrhage and intestinal villous atrophy.

Genetically, there are familial cases, cases in homozygous twins and association with the DR3 histocompatibility antigen. In Europe, associations with HLA B8, B18 and DQW2 have also been demonstrated, and in Japan with DR2 and DQW1. Some patients with GEM have a rare allotype of properdin factor B (BfF1) associated with HLA B18 and DR3.

The GEM and diabetes association would be linked to the common genetic susceptibility of these two diseases. The polymorphism of the gene TAP1, a gene encoding a transporter protein involved in antigen presentation, is increased in patients with GEM.

This gene is present in the region of the major HLA II histocompatibility complex.

Evolution and prognostic factors:

Renal survival is about 90% at 5 years and 65% at 10 years. The precise evaluation of the evolutionary potential of the disease must make it possible, at the initial stage and during the evolution, to assess the necessity of starting a treatment (assessment of the benefit / risk ratio).

The work of Schieppati is a prospective study on GEM performed in patients without treatment. In 100 patients with GEM, the probability of renal survival at 5 years was 88 ± 5% and at 10 years of 73 ± 7%. This study included 37 non-nephrotic patients and the mean proteinuria on the cohort was 5 + – 3.6 g 24 hours -1 . Sixty-two percent of women and 59% of men were partially or completely remission at 4 years of age. The only unfavorable prognostic factors were male and age over 50 years. At the time of diagnosis, the presence of nephrotic syndrome, arterial hypertension, proteinuria levels are not prognostic factors. Since the publication of this article and the advent of immunosuppressive treatments in GEM, in a decade, has the risk of a patient presenting with GEM been improved?

Pei et al. attempted to quantify the evolution risk using a model using the most important prognostic factors, namely proteinuria and its duration, creatinine clearance and slope 1 / serum creatinine as a function of time. This study included 184 patients diagnosed with normal renal function. On a follow-up average of 5.8 years, 26% of patients will develop renal insufficiency and 26% will go into remission. Fifty percent of patients received corticosteroids and 26 patients received an additional immunosuppressant. In the model studied for prognosis, immunosuppressive therapy had no influence on the incidence of renal insufficiency or on the percentage of remission. Proteinuria without a notion of its duration had little predictive value. A proteinuria value greater than 8 g 24 hours -1 for more than 6 months is associated with the highest risk of developing chronic renal insufficiency, with a probability of 66%. Conversely, even when it is greater than 8 g 24 hours -1 , proteinuria of duration less than 6 months has a negative predictive value of 88% chronic renal failure, reflecting the possibility of spontaneous remission regardless of the level of proteinuria . If proteinuria is between 6 and 8 g 24 hours -1 and lasts 9 months or more, the risk of kidney failure is 55%.

This model will benefit from a wide validation on three populations originating from Canada, Italy and Finland (363 patients). To predict the progression to chronic renal insufficiency (creatinine clearance < 60 ml min -1 ), the probability of progression takes the form:

= exponential Y / 1 + exponential Y with Y = [1,26 + (0,3 Å – mean proteinuria rate over the follow-up period in g / 24 h -1 ) – 0,3 Å ~ creatinine during follow-up in ml min -1 and month) – 0.05 Å creatinine clearance at entry in follow-up period].

This model is applicable for patients with a clearance greater than 60 ml min -1 at diagnosis. The model improves accuracy (sensitivity to specificity) compared to proteinuria alone. Compared to a proteinuria criterion greater than 3.5 gj -1 , the model improves specificity (89 versus 28%) but decreases the sensitivity (75 versus 92%). Compared to a criterion of proteinuria greater than 8 gj -1 for more than 6 months, the model improves sensitivity (83 versus 58%) but decreases specificity (86 versus 93%).

Other prognostic factors were also highlighted and confirmed in a meta-analysis of 1,248 patients who included prospective and retrospective studies. Of particular note is the excretion of immunoglobulin G (IgG), C5b-9 and b2 microglobulins, which are associated with a risk of progression to renal insufficiency with high predictive power.

Treatment:

SYMPTOMATIC TREATMENT:

Symptomatic treatment should be discussed regardless of the level of renal function and proteinuria.Immunosuppressive therapy has a place for patients with persistent nephrotic syndrome and renal function.

Optimal symptomatic treatment may include: ACE inhibitor or angiotensin II inhibitor or combination thereof and inhibitor of hydroxy-3-methyl glutaryl coenzyme A (HMG CoA) reductase.

In prospective trials, ACE decreases protein excretion by 25-50% in nephrotic patients.

Some patients, however, did not respond to treatment significantly, which could be an additional prognostic factor.Praga reported an antiproteinuric effect ranging from 0 to 85% among 11 patients. The response to ACE inhibitors may well be an independent prognostic factor unrelated to the degree of proteinuria or excretion of C5b-9. Reduction of proteinuria is a good marker for improvement of renal prognosis even if the direct demonstration of an improvement in the prognosis of GEM by ACE inhibitors has not yet been made.

The effect of angiotensin II inhibitors was less studied in the case of GEM than that of ACE inhibitors. The antiprotein effect is likely, by analogy with the effect demonstrated in diabetic glomerulopathy. Only one case has been reported to date with a reduction in proteinuria. Finally, given the value of the combination of angiotensin I inhibitor II, demonstrated by prospective studies in diabetic glomerulopathy (COOPERATE), this association deserves an evaluation in GEM and can already be part of symptomatic treatment at the outset or in case of persistence of nephrotic proteinuria under IEC.

The study in animals shows that hyperlipidemia accelerates the progression of renal insufficiency by inducing the equivalent of intraglomerular atherosclerotic lesions and accelerating interstitial fibrous lesions. In a prospective study of nine patients with nephrotic syndrome GEM, simvastatin, compared to diet alone, not only improved lipid balance, but also reduced proteinuria and a significant increase in albumin (26 at 46 gl -1 ). Despite this, the progression of renal insufficiency in the simvastatin group was comparable to that of the group with diet alone.

The effect of probucol has been demonstrated in animals with a proteinuria reduction of up to 85%. This effect involves the inhibition of lipid peroxidation and could block the alteration of the basal membrane, despite the formation of immune complexes on its surface. This effect was also found in human clinical practice in a series of 15 nephrotic patients resistant to immunosuppressive therapy (n = 7) and / or treated with IEC (n = 12) at a rate of 1 g per day for 3 months. A set of antioxidant treatments, also aimed at inhibiting the effect of lipid peroxidation, is being evaluated. The following scheme was proposed: vitamin C 1200 mg d -1 (to be adapted to renal function), tocopherol 600 IU j -1 , bcarotene 6 mg d -1 , selenium 60 μg d -1 , N-acetyl cysteine 200 mg d -1 . In a patient with a severe nephrotic syndrome resistant to immunosuppressive therapy, this combination resulted in a reduction in lipid peroxidation products, with urinary reduction of the C5b-9 attack complex and a parallel decrease in proteinuria.

Patients with GEM have a high thrombotic risk. In patients with nephrotic syndrome, the risk of thrombosis is estimated to be 35% (5-60% according to the studies) for renal vein thrombosis and 8-44% for thrombotic events from another location. In other words, the monthly incidence of clinical thrombosis is 0.5% for renal vein thrombosis and 1% for thrombosis at another site. These events are even more frequent in cases of GEM and severe hypoalbuminemia ( <20 gl -1 ). The probability of pulmonary embolism in untreated deep venous thrombosis is about 50%. Pulmonary embolism is a source of high mortality. On the other hand, anticoagulant therapy induces a hemorrhagic risk related to the intensity of the treatment and to risk factors:

– age over 65;

– history of intestinal haemorrhage;

– history of hemorrhage from another origin;

– severe comorbid conditions such as recent myocardial infarction, renal insufficiency, liver disease, cancer;

– supraventricular fibrillation.

The nephrotic syndrome induces greater variability in efficacy related to the variability of protein binding. In total, the incidence of bleeding events in patients with one or two risk factors is about 0.25% per month or 3% per year. This risk is 17% per year for patients with more than three risk factors. In decision analysis, prophylactic anticoagulant therapy is justified in nephrotic patients as long as they are affected by GEM and have no more than two risk factors for bleeding. Indeed, in these patients, the number of fatal pulmonary embolisms prevented by prophylactic anticoagulant therapy before any thrombotic event is greater than the number of haemorrhages induced by the same treatment.

Pentoxifylline (Torental t ) reduces the synthesis of TNF. It was tested in an open-label study of 1200 mg d -1 in 10 patients with nephrotic syndromic GEM, naïve of any treatment and with a duration of renal involvement ranging from 7 to 46 months. At 6 months of treatment, four were in partial remission of the nephrotic syndrome and five in complete remission.

Treatment has few side effects and could be advised even at this level of evidence (rank C recommendation).

IMMUNOSUPPRESSOR TREATMENTS:

Corticosteroids and alkylating agents (chlorambucil, cyclophosphamide):

Two meta-analyzes ensure quality recommendations (rank A). In the Hogan meta-analysis, seven prospective studies (including six randomized, but no placebo-controlled) were included. Five compared corticosteroids with no immunosuppressive therapy and three compared a combination of corticosteroids and an alkylating agent with no immunosuppressive therapy. These studies included only nephrotic patients. Complementary analysis included three studies that did not include a group without treatment or compared treatment other than an alkylating agent, and a final analysis included, in addition to these 10 studies, 22 retrospective series describing the course of treatment without corticosteroids or under an alkylating agent.

Thirty-three studies were excluded either because they did not differentiate between changes in the type of treatment or because they were only concerned with remission or relapse. Of these studies, 78% of the patients were nephrotic.Across the cohort of 1,189 patients, the renal survival rate is 0.86 to 5 years, 0.65 to 10 years and 0.59 to 15 years.The 5-year survival rate was identical in the seven prospective studies. At 5 years of age, in these same studies, the renal survival rate was 0.80 without treatment or subcorticoids and 0.99 with an alkylating agent. However, this difference did not reach the threshold of significance. The findings are similar across all 32 studies. According to prospective studies, the probability of remission is greater with the alkylating treatment compared to the absence of treatment and not modified by the corticoids.

The main conclusions of this meta-analysis are as follows: the renal prognosis is involved in the long term ( > 10 years); there is no beneficial renal effect of corticosteroids alone (either in terms of renal survival or remission of nephrotic syndrome); the alkylating agents have a beneficial effect in terms of the rate of remission of the nephrotic syndrome over a period of 24 to 36 months after the start of treatment, with no significant effect on renal survival.

Imperiale published identical results in a meta-analysis of five randomized prospective studies evaluating the benefit of the combination of corticosteroids and an alkylating agent, ie a total of 228 patients: alkylating agents increase the likelihood of complete remission by a factor of 4.6 (relative risk [RR] 2.2-9.3) and a response in terms of reduction in proteinuria of 2.3 (RR 1.7-3.2) with no effect which can be demonstrated on the preservation of renal function.

It should be noted that the meta-analyzes did not differentiate treatments that included chlorambucil or cyclophosphamide. Ponticelli found comparable efficacy in a prospective, randomized study of 97 patients treated with alternate monthly corticosteroid and alkylating agent (chlorambucil or cyclophosphamide) for 6 months.

At a lower level of evidence than these two meta-analyzes, there is a presumption in favor of preserving renal function with protocols involving an alkylating agent. The first argument is based on data from the Ponticelli team, which promotes these alternate treatments for 10 years of follow-up in nephrotic patients without renal impairment at the start of treatment (exact criterion: serum creatinine < 150 μmol l -1 ).

Of 81 patients randomized to receive immunosuppressive therapy or the only symptomatic treatment, the probability of living without dialysis at 10 years was 0.92 for immunosuppressed patients versus 0.60 for patients on symptomatic treatment; this difference is significant (p = 0.0038).

The second argument comes from retrospective studies that included patients with altered renal function. Reichert mentions a regression of creatininemia of 260 to 186 μmol l -1 in 10 patients, significant after 6 months of treatment, with an effect that seems to last up to 18 months and disappear at 3 years.

Branten has a similar effect under cyclophosphamide per os for 1 year, with a renal regression of 38% in median value in 39 patients (mean initial creatinine: 226 + – 108 μmol l -1 ). The improvement persisted at 36 months in 18 patients followed during this period. Approximately 40% of patients benefit from treatment, with progression to chronic terminal renal failure delayed by approximately 2.5 to 3 years. In a retrospective study of 39 patients with impaired renal function, 20 received symptomatic and 19 immunosuppressive therapy. The probability of renal survival at 4 years without dialysis was 55% for patients without immunosuppressant versus 90% with immunosuppressant. The severity factors at the time of the biopsy and the onset of renal insufficiency were comparable in both groups.

Finally, two retrospective studies were published after metaanalysis in patients without renal impairment. The results are mixed. In small groups of patients, they reported a reduction in the risk of end-stage renal disease, which was significant in the Polenakovik study and not significant in that of Stirling, which does not contradict the meta-analysis data.

Azathioprine:

Two old controlled studies have shown no benefit from treatment with azathioprine in combination with corticosteroids.More recent studies are retrospective. Ahuja found no favorable effects of azathioprinecorticoid combination in 38 patients compared with 20 controls without immunosuppressive therapy. Only in the study of Brown, in nephrotic patients with renal insufficiency, there was a benefit in association with a regression of renal involvement and a decrease in proteinuria. All these studies, however, are against the use of azathioprine.

Mycophenolate mofetil:

Mycophenolate mofetil (MMF) specifically inhibits the de novo synthesis route of purine bases by sparing the rescue pathway. This property allows, to a certain extent, a more specific action on the lymphocyte lines, which preferentially use this synthetic route while the other cell lines can resort to the rescue route. MMF has an antiproliferative action on lymphocytes, with a decreased response in the presence of mitogen and decreases the humoral response of B lymphocytes.

In the murine model of Heymann’s nephritis, MMF was able to prevent the onset of proteinuria, formation of glomerular deposits and interstitial infiltrate when administered within 4 weeks of injection of antigens. Beyond this time, the MMF is on the other hand ineffective.

Studies in humans are only open studies. Two authors reported their experience, which involved a total of 25 nephrotic patients. The effective dose is 1.5 to 2 g per 24 hours. All patients were at risk of corticosteroid + alkylating agent treatment, with the exception of three in whom MMF was prescribed from the outset. The results are a reduction in proteinuria in both studies, with no change in serum creatinine, with a median follow-up duration of 8 and 9 months respectively in the two studies. The response in terms of reduction of proteinuria is observed in both studies between 3 and 6 months. If no response is obtained at 6 months, treatment should be discontinued. The main side effects are leucopenia, gastrointestinal symptoms and viral infections by herpes virus.

The place of MMF in the treatment of GEM can not yet be established from the results of these two studies, which deserve to be supplemented by randomized trials comparing them to the corticosteroid and alkylating agents currently recommended in patients with high evolving risk. MMF could currently be proposed as second-line treatment if first-line therapy fails, as was the case in both open-label studies, or to limit corticosteroid therapy in relapse of nephrotic syndrome C).

Fludarabine:

Fludarabine is a nucleoside analogue of the purine bases which also has a more specific action on lymphocytes.

Boumpas reported seven cases of nephrotic GEM patients resistant to treatment with corticosteroids alone or with an alkylating agent or ciclosporin, treated for 6 months by monthly fludarabine cures. Proteinuria reduction was greater than 50% in five out of seven patients and not significant across the cohort. Renal function remained on average stable over 6 months. There is no longer-term follow-up study.

No recommendations can be derived from this single study.

Tetracoside (Synacthen):

Only an open-label study has so far been published, not allowing for tetracoside recommendation recommendation.The authors were primarily interested in the lipid-lowering effect and looked for a reduction in LDL and lipoprotein (a).

The study involved 14 patients with GEM with nephrotic syndrome. Four patients had previously been treated according to the Ponticelli regimen, seven were treated with corticosteroids alone and three had never been treated.They received tetracoside at a gradually increasing dose for 8 weeks until the optimal dose was reached, ie 1 mg three times a week. Subsequently, five patients with GEM with severe corticosteroid-resistant nephrotic syndrome were treated for 1 year at two injections of 1 mg per week. Levels of cholesterol, triglycerides, apolipoprotein B and lipoprotein (a) decreased significantly after 8 weeks of treatment, as did proteinuria (median reduction of 80%) and serum creatinine (median reduction of 28% ). Twelve patients out of 14 were put in partial remission of the nephrotic syndrome. Reduction of proteinuria and serum creatinine was maintained at 30 months for the five patients who continued treatment for 1 year. In patients treated for 8 weeks, discontinuation of therapy was followed by a further increase in serum creatinine and proteinuria within 1 month. The optimal dose proposed by the authors is 2 mg per week.

The results of this study appear to be spectacular as regards the number of patients put in partial remission and the rate of responses in terms of improvement of renal function. However, no recommendation from this open study can be formulated until future studies confirm or disprove this experiment.

Ciclosporin:

The randomized placebo-controlled study included 51 patients treated with corticosteroids at a dose of 0.15 mg kg -1 d -1 plus ciclosporin at a dose of 3.5 mg kg -1 d -1 or by corticosteroids at the same dose plus placebo. Seventy-five percent of patients (21/28) under ciclosporin had partial or complete remission without deterioration of renal function versus 22% in the placebo group. The percentage of remission at the end of the 78-week follow-up remained different between the groups (39 versus 13%). The rate of doubling of creatinine over the same period was similar in both groups (7 versus 9%). An initial increase in serum creatinine in the ciclosporin group was observed in nine patients versus three in the placebo group.

Reductions in ciclosporin doses were followed by an improvement in seven patients, while worsening continued in the remaining two patients despite discontinuation of treatment. The number of hypertensive patients and the severity of hypertension were greater under ciclosporin.

The first study carried out by the same team showed that in 17 patients, the rate of progression of renal insufficiency decreased (-2.05 vs -0.73 ml min -1 months -1 ) later study published in 2001.

The improvement of the nephrotic syndrome is therefore the major benefit of ciclosporin treatment, while the benefit in terms of preservation of renal function remains to be proved, especially since ciclosporin nephrotoxicity is superadded in the long term.

Rituximab:

Rituximab is a monoclonal antibody to the CD20 antigen present on the surface of B lymphocytes.

An inaugural trial tested rituximab in eight nephrotic patients with creatinine clearance greater than 20 ml min -1 .These patients had no remission from an average of 29 months, under full-dose IEC and statin, and had not received previous immunosuppressive therapy. They received a total of four weekly infusions of 375 mg m  2 rituximab. Over a period of 20 weeks, two patients entered full remission and three patients underwent partial remission (proteinuria <3.5 g 24 hours  1 ). At the end of the study, proteinuria had decreased by 62%. According to a complementary 1-year study, proteinuria was less than 0.5 g 24 hours -1 in two patients and less than 3.5 g 24 hours -1 in three patients.

Follow-up of these same patients showed persistence of benefit at 12 months, with a persistent 66% reduction in proteinuria compared to baseline. Renal function remained stable during the same period, but there was no analysis of renal function before and after treatment. No conclusion as to the preservation of renal function can currently be advanced.

The same team reported the predictive factors of response to rituximab as an abstract. Only histological factors have a predictive value, with a lower reduction of proteinuria in patients with high index of tubulointerstitial and vascular lesions.

Immunoglobulins:

A Japanese team reported its experience in 86 patients with GEM, followed at least 5 years, 30 of which were treated with open immunoglobulins. The treatment consisted of one to three immunoglobulin cures at 100 to 150 mg kg -1 d -1for 6 days. There were 68% and 70% of nephrotic patients respectively in the groups with and without Ig. Serum creatinine was less than 1.5 mg dl -1 in the majority of cases in both groups. In the non-Ig group, 39% of patients received corticosteroids associated with an alkylating agent and 30% of corticosteroids alone. Kaplan Meier’s 15-year survival analysis showed no benefit in terms of prolonging renal survival (18% renal loss in the Ig group versus 10% in the control group).