21.8.4 Membranous nephropathy 4928 An S. De Vriese

21.8.4 Membranous nephropathy 4928 An S. De Vriese and Fernando C. Fervenza

section 21  Disorders of the kidney and urinary tract 4928 FURTHER READING Braun N, et al. (2008). Immunosuppressive treatment for focal seg- mental glomerulosclerosis in adults. Cochrane Database Syst Rev, 3, CD003233. Brown EJ, Pollak MR, Barua M (2014). Genetic testing for nephrotic syndrome and FSGS in the era of next-​generation sequencing. Kidney Int, 85, 1030–​8. Iijima K, et al. (2014). Rituximab for childhood-​onset, complicated, frequently relapsing nephrotic syndrome or steroid-​dependent nephrotic syndrome:  a multicentre, double-​blind, randomised, placebo-​controlled trial. Lancet, 384, 1273–​81. Jalanko H (2009). Congenital nephrotic syndrome. Pediatr Nephrol, 24, 2121–​8. Kidney Disease:  Improving Global Outcomes (KDIGO) Glomerulonephritis Work Group (2012). KDIGO Clinical Practice Guideline for Glomerulonephritis. Kidney Int, 2 Suppl 2, 139–​274. Korbet SM (2012). Treatment of primary FSGS in adults. J Am Soc Nephrol, 23, 1769–​76. Maas RJ, Deegens JK, Wetzels JF (2014). Permeability factors in idio- pathic nephrotic syndrome: historical perspectives and lessons for the future. Nephrol Dial Transplant, 29, 2207–​16. McCarthy HJ, Saleem MA (2011). Genetics in clinical practice: neph- rotic and proteinuric syndromes. Nephron Exp Nephrol, 118, e1–​8. Shaloub (1974). Pathogenesis of lipid nephrosis: a disorder of T-​cell function. Lancet, 2, 556–​60. Siligato R, et al. (2018). Emerging therapeutic strategies for minimal change disease and focal and segmental glomerulosclerosis. Expert Opin Investig Drugs, 27, 839–79. Stokes MB, et al. (2004). Glomerular tip lesion: a distinct entity within the minimal change disease/​focal segmental glomerulosclerosis spectrum. Kidney Int, 65, 1690–​702. Zheng Q, et al. (2019). Comparative efficacy of 13 immunosuppressive agents for idiopathic membranous nephropathy in adults with neph- rotic syndrome: a systematic review and network meta-analysis. BMJ Open, 9(9), e030919. doi: 10.1136/bmjopen-2019-030919. 21.8.4  Membranous nephropathy An S. De Vriese and Fernando C. Fervenza ESSENTIALS Membranous nephropathy (MN) is the most common cause of nephrotic syndrome in Caucasians adults. It may also present with asymptomatic proteinuria. Its defining feature is the presence of subepithelial immune deposits, localized between the podocyte and the glomerular basement membrane. Aetiology—​primary MN (80% of cases) is caused in most cases by antibodies against the M-​type phospholipase A2 receptor (PLA2R). Secondary MN occurs in relation with drugs, malignancy, or auto- immune disease. Prognosis—​the clinical course of primary MN is variable: spontan- eous complete remission of proteinuria occurs in 20 to 30% and progressive kidney failure develops in 20 to 40% over 5 to 15 years. Patients with gross proteinuria (>8 g/​day) are at high risk of progres- sion, as are those with a high and rising anti-​PLA2R antibody level. Management—​patients at low risk of progression have an excellent long-​term prognosis and should be treated conservatively without immunosuppression. Patients at medium and high risk for progres- sion benefit from immunosuppression in addition to conservative treatment. Standard treatment regimens include corticosteroids with cyclophosphamide, calcineurin inhibitors (ciclosporin, tacrolimus) and rituximab. Introduction Membranous nephropathy (MN) is the most common cause of nephrotic syndrome in Caucasians adults. The defining feature is the presence of subepithelial immune deposits, localized between the podocyte and the glomerular basement membrane (GBM). These deposits and the subsequent formation of basement membrane ma- terial between and around them result in the thickened appearance of the GBM, hence the term ‘membranous’. Due to their localization on the subepithelial side of the GBM, the deposits do not result in in- filtration by inflammatory cells and subsequent glomerular inflam- mation. Thus, the term ‘nephropathy’ or ‘glomerulopathy’, rather than ‘glomerulonephritis’, is preferred. Primary MN, responsible for approximately 80% of cases, is an organ-specific autoimmune disease, in which circulating auto- antibodies bind to an autoantigen on the surface of the podocytes. Three major target antigens are now recognized: the M-type phospholipase A2 receptor 1 (PLA2R), the thrombospondin type 1 domain-containing 7A (THSD7A) and exostosin 1 (EXT1)/exostosin 2 (EXT2). Future research may reveal other—​as yet unknown—​ autoantibodies against components of the GBM in patients with MN. In approximately 20% of cases, MN occurs in relation with drugs, malignancy, or autoimmune disease, and the disease is categorized as secondary. Aetiology Primary membranous nephropathy The understanding of the autoimmune processes involved in primary MN has advanced substantially over the past few years. First was the discovery of antibodies against neutral endopeptidase (NEP) in rare cases of neonatal MN. Mothers with truncating mutations of the metallomembrane endopeptidase gene do not express NEP on cell membranes. NEP-​deficient mothers, who were immunized during pregnancy, transplacentally transfer nephritogenic antibodies against NEP to their children, causing MN in the newborns. Subsequently, antibodies to the PLA2R were found in 70 to 80% of patients with MN. The PLA2R is a member of the mannose receptor family. It was cloned based on its ability to bind secreted phospholipase A2 mol- ecules, but its exact cellular function is not fully understood. With the exception of Japanese patients who have a lower prevalence of PLA2R-​associated MN, the prevalence of these antibodies is con- sistent among ethnic patient populations around the world. More recently, 10% of the patients who are seronegative for anti-​PLA2R were found to have antibodies directed to the thrombospondin type-​1 domain-​containing 7A (THSD7A) protein. THSD7A is a transmem- brane protein expressed in many organs, including the kidney. Most recently, EXT1 and EXT2 were identified as target antigens in MN.

21.8.4  Membranous nephropathy 4929 Exostosins are glycosyltransferases responsible for the synthesis of the heparin sulfate backbone that add glycosaminoglycan residues to the core protein resulting in the generation of complex polysaccharides. The remaining 20% of cases may have disease caused by antibodies not yet identified or have secondary MN. Secondary membranous nephropathy The most frequent causes of secondary MN are listed in Table 21.8.4.1. A few cases of MN secondary to high levels of cir- culating antibovine serum albumin antibodies have been reported in children. In some geographic areas, infection is an important cause of secondary MN: malaria in Africa or hepatitis B in Asia. In older patients, solid tumours are the most common cause of sec- ondary MN. In some reports, there is a lag of months and some- times years from the time of the diagnosis of MN and detection of a tumour. Differentiating between primary MN and MN secondary to underlying malignancy is crucial because treatment approaches differ. Positive anti-​PLA2R antibodies have been reported in a few patients with MN associated with solid tumours, but resection of the tumour was not accompanied by remission of proteinuria, sug- gesting that the association of malignancy and MN was coincidental rather than causative. A  negative test for anti-​PLA2R antibodies emerged as the most important predictor of malignancy. Pathology In early MN, the glomeruli appear normal by light micros- copy. Increasing size and number of immune complexes in the subepithelial space produce thickening of the capillary walls. With time, new basement membrane is formed around the immune complexes. This produces the spikes and pinholes along the epithe- lial side of the GBM visible on the silver methenamine and peri- odic acid–​Schiff stains, as the deposits themselves do not stain (Fig. 21.8.4.1). In primary MN, proliferative features, including mesangial proliferation and endocapillary proliferation, are typically absent. Immunofluorescence microscopy shows diffuse and global bright granular (beaded) capillary wall staining for IgG, C3, kappa light chains, and lambda light chains, with negative staining for IgM, IgA, and C1q (Fig. 21.8.4.1). In some cases, the granular staining can be so dense that it may resemble a linear pattern. Staining for IgG sub- classes reveals predominance of IgG1 and IgG4. Electron microscopy shows numerous subepithelial deposits typically separated from each other by GBM material (Fig. 21.8.4.1). In primary MN, subendothelial deposits and mesangial deposits are not present, and the mesangium is usually unremarkable. Most cases of MN with active disease show extensive foot processes effacement. Some pathological features are strongly suggestive of a secondary cause of MN (Table 21.8.4.2). In MN caused by anti-​PLA2R antibodies, glomeruli may stain for PLA2R, in tight correlation with the presence of serum anti-​ PLA2R antibodies (Fig. 21.8.4.2). Some patients who are seronega- tive for anti-​PLA2R antibodies at the time of kidney biopsy show glomerular PLA2R positivity. Many of these cases may represent PLA2R-​associated disease that has already gone into immunological remission, leaving a historical clue to the immunological activity. A few seronegative, tissue-​positive cases may reflect very early dis- ease, with rapid clearance of the antibodies from the circulation and deposition in glomeruli due to the very high affinity between the anti-​PLA2R autoantibody and the abundant podocyte antigen. Clinical features Most patients with MN present with nephrotic syndrome, the re- mainder with asymptomatic proteinuria (usually <3.5 g/​24 h), com- monly found on a routine medical examination. At presentation, most are normotensive and have normal kidney function. The urine sedi- ment is generally bland, but microscopic haematuria is present in 30 to 40% of cases. The clinical features associated with nephrotic syndrome in MN can be severe, and ascites, pleural effusions, and pericardial ef- fusions may be present. Hyperlipidaemia, characterized by both an in- crease in total and low-​density lipoprotein cholesterol and a decrease in high-​density lipoprotein cholesterol, is common. Thromboembolic events frequently complicate the nephrotic syndrome in MN. The pre- cise mechanism of the hypercoagulable state in MN is unknown, but likely includes a combination of risk factors, such as a local decrease in perfusion pressure in the renal vein due to the low oncotic pressure, loss of anticlotting factors in the urine, increased hepatic production of clotting factors, as well as a genetic predisposition. Diagnosis The diagnosis of MN is made by kidney biopsy. Further diagnostic steps should be undertaken to differentiate primary from secondary MN. Careful medical history, physical exam, laboratory evaluation, and re- view of histological features are essential to identify potential secondary causes. Evaluation should include a complement profile, assays for antinuclear antibodies, rheumatoid factor, hepatitis B surface antigen and hepatitis C antibody, thyroid antibodies, and cryoglobulins. Testing for anti-​PLA2R and anti-THSD7A antibodies further aids the diag- nostic work-up. Initially, an indirect immunofluorescence test was used that provided semiquantitative titres. Currently, a higher-​throughput enzyme-​linked immunosorbent assay (ELISA) is commercially avail- able for these antibodies and offers a quantitative titre based on reference standards. Anti-EXT1/EXT2 antibodies have not yet been identified. Patients aged older than 60 years should undergo appropriate screening for underlying malignancy. There should be a low threshold for a CT scan of the thorax in patients with risk factors for lung cancer, upper and lower gastrointestinal endoscopy in anyone with abdominal or bowel symptoms of any sort, and mammography in women, particularly in Table 21.8.4.1  Secondary causes of MN Aetiology Examples Neoplasm Carcinomas (especially solid organ tumours of the lung, colon, breast, kidney, stomach, and prostate), Infections Malaria, hepatitis B and C, secondary or congenital syphilis, leprosy, filariasis Drugs Penicillamine, gold, captopril, mercury, nonsteroidal anti-​inflammatory drugs Immunological Systemic lupus erythematosus, mixed connective tissue disease, thyroiditis, dermatitis herpetiformis, rheumatoid arthritis Miscellaneous Sickle cell anaemia; graft-​versus-​host disease, sarcoidosis Bovine serum albumin In children

section 21  Disorders of the kidney and urinary tract 4930 (a) (b) (c) (d) Fig. 21.8.4.1  (a, b) Light microscopy showing thickened glomerular basement with pin holes and spikes ((a) periodic acid Schiff stain, (b) silver methenamine stain; both magnification ×40). (c) Immunofluorescence microscopy showing granular IgG along the capillary walls (magnification ×40). (d) Electron microscopy showing numerous subepithelial deposits (black arrows). Note basement membrane material between the deposits (white arrows) forming the spikes (magnification ×6800). Courtesy of Dr. Sanjeev Sethi, Division of Anatomic Pathology, Mayo Clinic, Rochester, MN. Table 21.8.4.2  Pathological clues to a secondary cause of MN Light microscopy Mesangial or endocapillary proliferation Immunofluorescence microscopy Full-​house pattern of Ig staining (IgG/​ IgA/​IgM) and C1q staining IgG subclass staining IgG1, IgG2, IgG3 → class V lupus nephritis Absence of IgG4 → MN secondary to malignancy. Electron microscopy Electron-​dense deposits located in the subendothelium and/​or mesangium or along the tubular basement membrane and vessel walls, presence of endothelial tubuloreticular inclusions Only few superficially scattered subepithelial deposits → malignancy-​ or drug-​associated MN Fig. 21.8.4.2  Membranous nephropathy showing positive granular staining for PLA2R along the glomerular basement membranes (magnification ×40). Courtesy of Dr. Mariam P. Alexander, Division of Anatomic Pathology, Mayo Clinic, Rochester, MN.

21.8.4  Membranous nephropathy 4931 patients who test negative for anti-​PLA2R antibodies. In patients with preserved kidney function and no evidence of secondary causes, a posi- tive PLA2R test highly predicts a tissue diagnosis of PLA2R-associated MN. In this subgroup of patients, a renal biopsy may not be necessary to make treatment decisions, especially for patients at high risk of compli- cations or in whom a renal biopsy is contraindicated. Natural history The clinical course of MN is variable. Spontaneous complete remis- sion of proteinuria occurs in 20 to 30% of patients with primary MN. Progressive kidney failure develops in 20 to 40% of cases over 5 to 15 years of observation. In the remaining patients, mild to moderate pro- teinuria persists. A complete remission and a lower relapse rate are more common in patients with persistent low-​grade (subnephrotic) protein- uria and in females. In contrast, male sex, age greater than 50 years, high levels of proteinuria (>10 g/​day), abnormal kidney function at presenta- tion, and tubulointerstitial disease, including focal and segmental lesions on biopsy, have all been associated with a lower renal survival. In a few cases, a rapid change in either the degree of proteinuria or the rate of loss of renal function may occur. This event should raise the suspicion of a superimposed condition, for example, acute renal vein thrombosis, superimposed acute interstitial nephritis, or cres- centic glomerulonephritis (antineutrophil cytoplasmic antibody-​ associated vasculitis or anti-​GBM disease). Anti-​PLA2R levels and disease activity Levels of circulating anti-​PLA2R antibodies correlate with clinical status, that is, the disappearance of the antibody is associated with remission of proteinuria, while reappearance of the antibody may herald a relapse of the nephrotic syndrome. Spontaneous remission is more likely in patients with low or moderate anti-​PLA2R antibody titres, but is rare in patients with high antibody levels. High levels of anti-​PLA2R antibodies are associated with progression of protein- uria and declining kidney function over time. Further, changes in antibody levels precede changes in proteinuria, independent of the type of immunosuppressive treatment. Serial measurement of anti-​ PLA2R antibodies levels may help in monitoring disease activity and response to immunosuppression. Preliminary data show that cyto- toxic agents, calcineurin inhibitors, mycophenolate mofetil (MMF), adrenocorticotropic hormone (ACTH), and rituximab all can induce depletion of circulating anti-​PLA2R antibodies. Finally, patients with high anti-​PLA2R antibody levels are more prone to relapse. Risk prediction and timing of therapy The principal challenge in the treatment of MN is to identify those patients at risk for progression. Until recently, the best prediction of renal outcome was a semiquantitative model using proteinuria and level of kidney func- tion. Patients who present with proteinuria up to 4 g/​24 h and stable kidney function over a 6-​month observation period are classified as being at low risk of progression. Patients with normal kidney function and stable creatinine clearance during a 6-​month observation period, but persistent proteinuria of between 4 and 8 g/​24 h, have a 55% prob- ability of developing chronic kidney injury, and are classified as being at medium risk of progression. Patients with persistent proteinuria greater than 8 g/​24 h, independent of the degree of kidney function, have a 66 to 80% probability of progression to chronic renal failure within 10 years, and are classified as being at high risk of progression. Patients at low risk of progression have an excellent long-​term prog- nosis and should be treated conservatively without immunosuppres- sion. Patients at medium risk and at high risk for progression benefit from immunosuppression in addition to conservative treatment. The commercial availability of an ELISA that quantifies anti-​ PLA2R antibody titres profoundly affects risk prediction in pa- tients with anti-​PLA2R antibody-​positive MN. As discussed earlier, emerging data demonstrate that the titres are commensurate with dis- ease activity, hence the prediction of the renal outcome model based on proteinuria will need to be reconciled with the new information obtained by monitoring of anti-​PL2R antibody titres. For example, a patient with very high anti-​PLA2R antibody levels is unlikely to go into spontaneous remission and should be considered for early im- munosuppressive therapy, even if proteinuria is less than 10 g/​24 h. Treatment Conservative therapy Conservative treatment involves dietary sodium restriction to less than 4 g/​day, moderation of dietary protein intake (0.8 to 1 g/​kg per day), and control of blood pressure, hyperlipidaemia, and oedema, and applies to all patients. Drugs that block the angiotensin II system are preferred because of their antiproteinuric effect beyond that ex- pected by their antihypertensive action. However, the evidence that angiotensin II inhibition is beneficial in MN is weak and largely in- ferential. The degree of renal protection is related to the degree of proteinuria reduction: if proteinuria does not decrease substantially, the beneficial influence is attenuated. In patients with MN, the antiproteinuric effect is modest (<30% decrease) and is more signifi- cant in patients with lower levels of proteinuria. Whereas patients with diabetic nephropathy are generally severely hypertensive and require multiple antihypertensive drugs, patients with MN are usu- ally normotensive, precluding the use of high doses of angiotensin-​ converting enzyme inhibitors/​angiotensin II receptor blockers. Thus, these drugs may not offer the same degree of renal protection in patients with MN as in diabetic nephropathy. In fact, a multi- variate analysis on the use of angiotensin-​converting enzyme in- hibitors/​angiotensin II receptor blockers in MN did not show an independent protective effect. If a significant antiproteinuric re- sponse will occur, it is usually seen within 2 months of initiation of angiotensin II blockade. The aim is to reduce proteinuria as close as possible to less than 0.3 g/​24 h. This goal is unrealistic in patients with proteinuria greater than 5 g/​24 h, even at high drug doses. Hyperlipidaemia increases the risk for cardiovascular disease in patients with nephrotic syndrome and we advocate the use of statins in patients with MN and persistent proteinuria. There is no consensus on when to start prophylactic anticoagulation in patients with MN. A recent study showed that in patients with MN, a serum albumin level less than 28 g/​litre was the most significant independent pre- dictor of venous thromboembolism. A positive familial or personal history of a thrombotic event, prolonged serum albumin levels less

section 21  Disorders of the kidney and urinary tract 4932 than 20 g/​litre, immobilization, and obesity should prompt consid- eration of prophylactic anticoagulation. Immunosuppressive therapy Corticosteroids Corticosteroid monotherapy is ineffective in preventing disease pro- gression and monotherapy with these drugs should not be used in primary MN. Corticosteroids with alkylating agents An initial study by Ponticelli and colleagues evaluated whether a regimen consisting of 1 g of intravenous methylprednisolone on the first 3 days of months 1, 3, and 5, followed by 27 days of oral prednisone at 0.5 mg/​kg, and alternating in months 2, 4, and 6 with chlorambucil at 0.2 mg/​kg per day for 30 days, was beneficial in patients with MN at moderate risk of progression. The prob- ability of achieving complete or partial remission was significantly higher in the group treated with immunosuppression versus con- servative therapy (83 vs 38% respectively). About 40% of conserva- tively treated patients reached endstage renal disease after 10 years, compared with only 8% of patients treated with steroids/​cytotoxic agents. The original regimen appeared remarkably safe and all ad- verse events were reversed after withdrawal of the drugs. In a sub- sequent study, the same regimen but containing cyclophosphamide (2.5 mg/​kg per day) rather than chlorambucil was found to be equally effective, better tolerated, and safer. Cyclophosphamide has become the preferred cytotoxic drug since. The beneficial effects of the studies by Ponticelli and colleagues have been confirmed by an Indian randomized controlled trial on 93 patients. The most frequent side effects of cyclophosphamide are leuko- penia and infection. We recommend that blood cell count should be checked weekly and the dose of cyclophosphamide be stopped if the total white cell count falls below 3000/​mm3. The target is to obtain lymphocyte counts below 700/mm3. Since cyclophosphamide can cause azoospermia and ovarian failure, patients at reproductive age are encouraged to undergo fertility consultation prior to starting therapy (sperm cryopreservation and ovarian protection therapy, e.g. leuprolide acetate). Cyclophosphamide accumulates with age and decreased kidney function and we recommend dose adjustments for age (>60 years, reduce dose by 25%) and serum creatinine (>220 μmol/​L (2.5 mg/​dl) reduce dose by 25%). A main concern with cyto- toxic drugs is the long-​term risk of neoplasia in patients receiving cyclophosphamide when the cumulative dose exceeds 36 g. Corticosteroids with mycophenolate mofetil MMF (2 g/​day for 12  months) combined with intravenous methylprednisolone (1 g the first 3  days of months 1, 3, and 5, followed by alternate-​day prednisone at 0.5 mg/​kg per 48 h for 6 months) may result in response rates similar to steroid therapy for 6 months combined with cyclophosphamide (1.5 mg/​kg per day) for 12 months. However, 75% of the patients treated with MMF relapsed within 2 years of treatment cessation. Monotherapy with MMF ap- pears ineffective in primary MN. Calcineurin inhibitors Therapy with a calcineurin inhibitor, that is, ciclosporin (3.5 to 5 mg/​ kg per day) or tacrolimus (0.05 to 0.1 mg/​kg per day), results in similar rates of remission as combined steroid/​cytotoxic therapy in patients at medium risk of progression. Relapses are common following dis- continuation and treatment should be maintained for at least 1 year. These drugs require monitoring for nephrotoxicity, hypertension, and development of diabetes. A recent multicentre randomized clin- ical trial found that ciclosporin did not prevent decline of kidney function in patients with primary MN and was associated with an excess of side effects as compared to placebo. However, a number of concerns have been raised and the study may have underestimated the efficacy of ciclosporin, one of the criticisms being the surrogate renal endpoint used (i.e. a mere 20% drop in estimated glomerular fil- tration rate). Too many variables, such as lowering of blood pressure, use of diuretics, and change in kidney creatinine handling during nephrotic syndrome, may contribute to slight changes in serum cre- atinine values, as well as the high ciclosporin starting dose of 5 mg/​ kg, which may have induced a sudden drop of estimated glomerular filtration rate and result in treatment failure as per protocol. Adrenocorticotropic hormone The use of ACTH, either under synthetic form (Synacthen, 1 mg subcutaneously twice weekly) or as a natural, highly purified gel for- mulation (H.P. Acthar gel, 80U subcutaneously twice weekly) has demonstrated promising results in primary MN. Rituximab Experimental data suggest that B cells are involved in the pathogen- esis of MN. Rituximab is a chimeric monoclonal antibody that acts through ligation with the membrane receptor CD20 of B cells, and inhibits their activation, proliferation, differentiation, and immuno- globulin secretion. In contrast to cyclophosphamide, which has striking but nonselective effects on B-​cell function, rituximab offers a more targeted approach to B-​cell depletion. In patients with MN and persistent nephrotic syndrome, rituximab given as once-​weekly infusions for 4 weeks of 375 mg/​m2 (the ‘lymphoma’ regimen) significantly decreases proteinuria. The response in proteinuria is gradual and sustained and may extend over a period of 2 years. Total B-​cell counts start to recover at 3 months, which is faster than in patients with other autoimmune nonproteinuric conditions, but there is no correl- ation between rituximab levels, degree of proteinuria, or response to the drug. The ‘rheumatology’ regimen, consisting of 1 g intravenously, given twice 2 weeks apart, was found to be as effective as the regimen with four weekly infusions. Finally, a B-​cell titrated protocol using a single dose of 1 g rituximab was shown to be equally successful as the four-​dose protocol but at a lower cost. Rituximab may also allow withdrawal of calcineurin inhibitors in dependent patients. A recent multicentre randomized con- trolled trial of rituximab versus ciclosporin in patients with severe MN (MENTOR) revealed that rituximab is not inferior to ciclosporin in inducing complete or partial remission of proteinuria and is superior in maintaining long-term remission of proteinuria. In summary, rituximab appears effective in inducing remission of proteinuria in many patients with MN, not only as initial treatment but also in those refractory to other treatments. The few short-​term side effects and assurance of compliance are advantages over more conventional immunosuppressive regimens, although some concerns about the long-​term risks of rare and fatal com- plications such as progressive multifocal leukoencephalopathy remain. Prophylaxis Corticosteroids/​cytotoxic agents and rituximab increase the risk of Pneumocystis jirovecii pneumonia and prophylaxis (usually with