The Kidneys and Lupus
IntroductionThe overall survival of patients with lupus nephritis has improved considerably over the last few decades; from less than 50 % survival at 5 years in the 1960s to over 80% survival at 20 years in the 1990s. This improved survival is due to a combination of factors: including the wider use of corticosteroids and immunosuppressants and the availability of more effective anti-hypertensive drugs, antibiotics, renal dialysis and transplantation. Early deaths from extra-renal lupus and infection are now uncommon but, instead, renal failure and cardiovascular disease have emerged as important determinants of morbidity and mortality.
PathogenesisThe pathogenesis of lupus in general, and lupus nephritis in particular, is complex and multifactorial.
See page - Genetics
Clinical phenotypesThe American College of Rheumatology has published criteria for the classification of the disease.
See - Diagnosis of Lupus - Specific Symptoms
In many patients, lupus is characterised by intermittent flares of constitutional symptoms, without evidence of major organ involvement. Involvement of major organ systems may occur in the heart, lungs, kidneys or central nervous system (Table 1). These are responsible for most of the mortality and morbidity in this disorder. These manifestations may develop asynchronously over time and this can lead to delays in establishing the diagnosis.
Lupus NephritisClinically apparent nephritis develops in about 40-75% of patients with lupus and the kidney is the organ most commonly affected. The diagnosis is made by the presence of proteinuria plus or minus haematuria. All patients with lupus should have their urine tested for protein and blood at each clinic visit. Nephritis typically develops early in the course of lupus and, in most patients, lupus nephritis will have appeared within 5 years of diagnosis.
Three major patterns of lupus nephritis have been defined, based on renal histology: focal proliferative, diffuse proliferative and membranous glomerulonephritis. This classification has subsequently been expanded by a World Health Organisation (WHO) committee (Table 2). Patients with minimal changes or mesangial glomerulonephritis (WHO class I and II lesions) usually have an inherently low rate of progressive renal failure. Patients with membranous nephropathy (WHO class V) have an intermediate prognosis for renal function.
By contrast, patients with focal or diffuse proliferative glomerulonephritis (WHO class III and IV) have a high risk of progressive renal failure.This classification was revised in 2003 by the International Society of Nephrology and the Renal Pathology Society.The histological classes range from normal by light microscopy (class 1) to advanced sclerotic nephritis (class VI).
Renal PresentationIn almost all cases, lupus nephritis develops in patients with evident systemic lupus erythematosus and extra-renal symptoms such as a rash, arthralgia, Raynaud's phenomenon and pleuro-pericarditis predominate. Rarely, renal disease may be the presenting feature of lupus with a small number of patients subsequently, after a period of months to years, developing extra-renal signs of lupus.
Proteinuria is a common finding and is often accompanied by a nephrotic syndrome.This is usually accompanied by various degrees of impairment of renal function. Microscopic haematuria is common but an acute nephritic syndrome is not. A few patients present with a rapidly progressive glomerulonephritis which may be severe enough to lead to acute renal failure. In such patients, a diffuse crescentic nephritis with intra-capillary glomerular thrombi is often seen.
Investigations in Lupus (Table 3)
Laboratory testsPatients with active lupus commonly have a leucopenia, lymphopenia and thrombocytopenia, often in association with an anaemia of the normochromic normocytic type. The anaemia may result from immune-mediated haemolysis with a positive Coombs’ test. An elevated serum creatinine concentration indicates underlying renal involvement, this is an adverse prognostic factor predicting subsequent development of renal failure.The presence of haematuria and proteinuria, particularly if accompanied by active urinary sediment with red cells and casts, is suggestive of active renal disease.
Patients with lupus, in common with those with other autoimmune rheumatic diseases, have raised polyclonal immunoglobulin concentrations but a normal acute phase response providing there is no additional infection.The erythrocyte sedimentation rate (ESR), which is affected by serum proteins and acute phase reactants, has come to be regarded as unreliable in diagnosis. It can, however, be a useful non-specific marker for distinguishing between active and inactive lupus. It may be normal even in the presence of major organ involvement and, conversely, may remain raised in remission. C reactive protein, the classical acute phase protein, is seldom raised in active disease unless the exacerbation is accompanied by serositis or infection. Neither the ESR nor CRP is specific for active lupus so their usefulness in monitoring disease activity has to be interpreted in this light.
ImmunologyPatients with lupus almost invariably express antibodies to components of the cell nucleus (ANA). A fluorescent antinuclear test is positive in greater than 95% of patients. In this test, various staining patterns (homogeneous, speckled, rim, nucleolar) can be demonstrated depending on the content of different autoantibodies in the serum. A positive fluorescent ANA test is useful because of its sensitivity, although it lacks specificity and a positive fluorescent ANA test can be found in other connective tissue diseases. More specific, but less sensitive, findings include anti-double-stranded DNA (anti-dsDNA) and anti-Sm autoantibodies. Anti-dsDNA antibodies bind the helical backbone of native DNA, whereas anti- Sm antibodies bind to proteins on an RNA-protein complex termed snRNP. Antibodies to double stranded DNA have the highest specificity for SLE but are only present in about 50% of patients with lupus. The interpretation of antidsDNA antibodies is also complicated by the lack of standardisation and various assays for anti-DNA antibodies are currently available: the most commonly used are the ELISA, Crithidia immunofluorescence assay and the Farr radiobinding assay.
These assays differ in the source of DNA and the physical chemical properties of the anti-dsDNA antibodies detected. In general, anti-dsDNA antibody levels reflect disease activity, particularly if accompanied by falling complement levels. Antinuclear and anti-dsDNA antibody levels have been less consistently related to features of active glomerulonephritis.
Patients with lupus will often have a false positive response to serological tests for syphilis such as the Venereal Disease Research Laboratory (VDRL) and rapid plasma reagin (RPR) tests because of antibodies to cardiolipin or other phospholipids. Often, antibodies to phospholipids will also interfere with clotting tests resulting in prolongation of the partial thromboplastin time. Anticardiolipin antibodies and/or the lupus anticoagulant are also found in patients with the anti-phospholipid syndrome and are associated with arterial thrombosis and miscarriages.
Reduced serum complement concentrations are useful in diagnosis and in assessing disease activity. Many patients with lupus have activation of the complement cascade with consumption of C1q, C4 and C3. Persistent C3 or CH50 complement depression has been associated with progression of renal disease in some patients.
Although changes in laboratory tests are extremely useful indicators of disease activity, most doctors do not base therapeutic decisions solely upon them because their sensitivity is less than absolute.Disease activity assessment in lupus must use the additional information provided by the clinical features.
Role of renal biopsyA renal biopsy is justified when there is evidence of glomerular disease, particularly if this is accompanied by extra-renal features of disease. Glomerular disease is likely if there is proteinuria; (>)200mg/24 hour; or protein/creatinine ratio (>)100mg/mmol, and/or haematuria (>)10 dysmorphic red blood cells per high power field, or casts of red and white blood cells), with or without renal insufficiency. Histology allows an assessment of disease activity and provides a basis for therapeutic options as well as providing prognostic information.
Monitoring disease activityDisease monitoring should always include an assessment of clinical symptoms. Objective laboratory measurements of anaemia, thrombocytopenia, leucopenia, lymphopenia, serum creatinine and estimated glomerular filtration rate (modification of diet in renal failure (MDRD) equation) is useful for monitoring lupus activity. The levels of complement and auto-antibodies, including antinuclear antibodies, anti-dsDNA antibodies may be useful markers of disease activity. Urinalysis should be performed at each visit. The persistence/emergence of haematuria and/or proteinuria suggests active renal disease.
Approach to the therapy of lupus treatmentThe majority of patients with lupus are successfully managed with non-steroidal anti-inflammatory agents, hydroxychloroquine and low dose corticosteroids. Drug regimens are increased in response to flares and gradually tapered during periods of remission. High dose corticosteroids and immunosuppressive drugs are reserved for patients with life threatening manifestations including severe lupus nephritis, central nervous system, cardiopulmonary disease or haematological abnormalities such as thrombocytopenia.
Treatment of lupus nephritisThere are several considerations in the approach to the treatment of patients with lupus nephritis. The first is based on the histological severity of the renal lesion.The second is based on the severity of the clinical presentation.The third consideration is the choice of therapy for inducing remission of acute disease and for maintaining remission and treating relapses.The heterogeneity of the clinical course of lupus nephritis and the relatively few randomized controlled trials make choice of treatment difficult and there is still substantial disagreement on the optimum treatment of lupus nephritis.
Mesangial Proliferative Glomerulonephritis (WHO class II)Most such patients present with proteinuria and microscopic haematuria, often with little in the way of renal impairment.There are no controlled trials to guide treatment. Patients are treated with corticosteroids (0.1-0.5mg/kg prednisolone/day tapering over months) in the hope that this will prevent progression to a more severe glomerulonephritis although this cannot be assured.
Membranous Nephropathy (WHO class V)In patients with lupus nephritis, the frequency of membranous nephropathy is approximately 12% when the definition of the renal histology is confined to pure membranous nephropathy or with mild mesangial hypercellularity, expansion and scattered deposits (WHO classes Va and Vb).The frequency increases to approximately 26% when there is in addition a focal segmental proliferative (WHO class Vc) or diffuse proliferative glomerulonephritis (WHO class Vd). The clinical presentation is with proteinuria and, in about 50% of cases, a nephrotic syndrome. Patients with WHO class Va and Vb lesions have a low rate of progressive renal failure whilst patients with a WHO class Vc or Vd lesions have higher risk of progressive renal failure which is comparable to that of patients with a proliferative glomerulonephritis. Here, again, there are no controlled trials of treatment and so there is no consensus on treatment.
In some studies, patients with WHO class Va and Vb have been treated with prednisolone and a smaller proportion also received methylprednisolone pulses and oral cyclophosphamide or azathioprine. By contrast, most patients with WHO class Vc and Vd have been treated with cyclophosphamide or azathioprine in addition to prednisolone.With these approaches to treatment, the 10-year survival free of death and renal failure in WHO class Va and Vb was 72-92% and in WHO class Vc and Vd was 35-81%. It seems reasonable to treat patients with membranous nephropathy and proliferative glomerulonephritis with cyclophosphamide as well as prednisolone because of the proven efficacy of this regime in patients with proliferative lupus glomerulonephritis. Patients with pure lupus membranous nephropathy with or without minor mesangial proliferation may be treated with prednisolone and consideration should be given to adding in azathioprine, cyclosporin or mycophenolate mofetil as a corticosteroid-sparing agent.
Focal and Diffuse Lupus Proliferative Glomerulonephritis (WHO class III and IV)As the prognosis of patients with these types of lupus nephritis was much poorer than those of patients with mesangial proliferative glomerulonephritis and membranous nephropathy, these patients have been the focus of most of the clinical trials of treatment. In patients with severe focal proliferative or diffuse proliferative glomerulonephritis, corticosteroids alone will reduce the extra-renal manifestations but are less efficient at preserving renal function. There is now good evidence that the addition of cyclophosphamide to prednisolone confers benefit when compared with patients treated with prednisolone alone. The evidence that azathioprine confers such benefit is less good although this agent may have a role in maintaining remission.
A meta-analysis of randomised controlled studies has shown that when compared with prednisolone on its own, cyclophosphamide and prednisolone reduced the risk of doubling of the serum creatinine (RR 0.59; 95%CI 0.4-0.88) whilst azathioprine did not (RR 0.98; 95% CI 0.36-2.68). Neither drug reduced the risk of developing end stage renal failure although further meta-analysis shows that cyclophosphamide does so. Azathioprine reduced the risk of death (RR 0.60; 95% CI 0.36-0.99) whilst cyclophosphamide did not (RR 0.98; 95%CI 0.53-1.82). Cyclophosphamide increased the risk of sustained amenorrhoea (RR2.18; 95%CI: 1.10-4.34).
A series of clinical trials from the National Institutes of Health (NIH) provided evidence of the effectiveness of intermittent intravenous cyclophosphamide together with oral prednisolone in preserving renal function in patients with severe lupus nephritis. This regime is preferable to continuous oral cyclophosphamide as it typically leads to a lower cumulative dose of cyclophosphamide.
From the NIH data, monthly pulse cyclophosphamide (0.5- 1.0 g/m2) adjusted for the glomerular filtration rate and leucocyte count at 10-14 days is given monthly for the first 6 months, then quarterly for 18-24 months. The longer course of cyclophosphamide has been associated with fewer relapses than a shorter 6 month course but is associated with greater gonadal toxicity. Preliminary data with pulse oral cyclophosphamide has shown encouraging results and, if validated, will minimize the inconvenience associated with intravenous therapy. To reduce the bladder toxicity of intravenous cyclophosphamide, patients should be hydrated either with oral or intravenous fluid and 2-mercaptoethane sulfonate sodium (mesna) given concomitantly. Prednisolone is given in conjunction with the cyclophosphamide at an initial dose of (0.5-1 mg/Kg/day) for 6-8 weeks with gradual tapering to minimize toxicity.This regime has been reported in detail and checks must be made prior to the use of such immunosuppressive regime.
The clinical benefit of mycophenolate was first confirmed in a number of small uncontrolled studies and then by larger randomized controlled trials. A conclusion from these studies may be that mycophenolate is probably as effective as and less toxic than cyclophosphamide in patients with new onset mild to moderate lupus nephritis.
Another study assessed the role of mycophenolate, azathioprine and quarterly intravenous cyclophosphamide as maintenance therapy in three groups of lupus nephritis patients (mainly classes III and IV) who all received induction with monthly intravenous cyclophosphamide.The cumulative probability of remaining relapse free was higher in the mycophenolate (78%) and azathioprine (58%) compared to the cyclophosphamide group (43%) after a median treatment duration of 29, 30 and 25 months. The event-free survival rates for the composite end-points of death and chronic renal failure were higher in the mycophenolate and azathioprine groups than the cyclophosphamide group. Hospitalisation, amenorrhoea and infections were lower in the mycophenolate or azathioprine groups, compared to cyclophosphamide.
In summary, further large adequately powered randomised controlled trials and longer term follow up is necessary to establish the role of mycophenolate in induction treatment of lupus nephritis and also as a maintenance therapy.
Drug toxicitiesThe various drugs used in the treatment of lupus are discussed elsewhere on this website.
See page - Drug Therapy of Lupus
These additional notes are applicable to lupus nephritis.
MycophenolateMycophenolate is widely used in the field of human organ transplantation. The active metabolite of mycophenolate is an inhibitor of purine synthesis. It blocks the proliferation of activated T and B lymphocytes and decreases antibody formation. Mycophenolate is teratogenic in animal studies and is, therefore, contraindicated during pregnancy. In general,mycophenolate is well tolerated and most of the adverse events respond to a reduction in dose. Gastrointestinal intolerance, particularly nausea and mild to moderate diarrhoea, occur in up to 10-40% of patients.There is an increased risk of infection which can be associated with leucopenia and lymphopenia.
Other therapeutic options
General measuresAs with other proteinuric renal diseases, angiotensin converting enzyme inhibitors or angiotensin II receptor blockers are recommended in patients with lupus nephritis. Hyperlipidaemia should be treated with statins in view of the increased risk of vascular disease in patients with lupus. Finally, bone protection in the form of calcium and vitamin D supplements in patients on long term steroids should be used. Biphosphonates are contraindicated in women of childbearing age and these drugs should be avoided in female patients with lupus.
PlasmapheresisSeveral studies have examined the role of plasmapheresis in the treatment of patients with SLE and lupus nephritis. Although plasmapheresis was well tolerated with few adverse effects, impact on renal function was disappointing. The controlled trials have either shown slight but insignificant benefit, or no benefit of plasmapheresis. Since removal of autoantibodies leads to a compensatory enhanced production of autoantibodies by pathogenic B-cell clones, the concept of synchronizing plasmapheresis with subsequent pulse cyclophosphamide to target proliferating B-cell clones was introduced. Initial studies have shown good remission rates in patients with lupus nephritis, but patient numbers were small and the dose of cyclophosphamide used was high (1.2-1.4 g/m2). The toxicity was too high to recommend this approach.
Intravenous immunoglobulinUncontrolled studies have shown a temporary benefit in lupus patients from the infusion of high doses of intravenous immunoglobulin. Currently, the data on the use of intravenous immunoglobulin in the treatment of lupus nephritis is limited and, as such, this treatment cannot be recommended.
CyclosporinSeveral studies have examined the effectiveness of cyclosporin in the treatment of lupus nephritis. None of these studies were controlled and it is difficult to discern whether cyclosporin was of any benefit in lupus nephritis. The nephrotoxicity of cyclosporin is a major problem and, pending randomized controlled studies comparing this drug with other immunosuppressive agents, it cannot be recommended for use in lupus nephritis.
MethotrexateMethotrexate may be useful as a steroid-sparing agent in lupus with arthritis and serositis, and may have potential benefits in mild nephritis. Methotrexate is, however, renally excreted and cannot be used in patients with renal impairment.
Androgens or anti-estrogen therapyThe observations that lupus is a disease predominantly affecting young women with a tendency to flare during pregnancy or with oral contraceptive administration, suggests that hormonal factors may be determinants of pathogenesis or severity.Attempts in men to control lupus by manipulating estrogen and testosterone levels have had only modest success. Despite in vitro data, there is little evidence to support the use of androgens or anti-estrogen in lupus nephritis.
Monoclonal AntibodiesSeveral studies have examined the use of monoclonal antibody therapy in patients with lupus nephritis. CD5 is a molecule on the surface of T cells and a subpopulation of B cells, and the use of anti-CD5 ricin A chain immuno-conjugate (CD% PLUS) led to an improvement in urine and laboratory parameters in 5 of 8 patients with glomerulonephritis. This approach in clinical practice remains to be validated.
Inhibition of co-stimulationCD40 binding to CD40 ligand is one of the most important co-stimulatory signals on B cells inducing activation and proliferation. The first open-label study focused on 30 patients with lupus nephritis and showed improvement in serology. This study was halted because of unexpected thromboembolic events. A second double-blind placebo controlled trial of 85 patients with mild to moderate lupus failed to show clinical efficacy over placebo.
Alternative co-stimulatory targets in lupus include the CD2 and CTLA4 receptors and their B-cell co-ligands B&-1 and B7-2.The T cell antigen CTLA-4 linked to murine IgGg2a (CTLA4Ig) has been shown to block autoantibody production and prolong life span in NZB/NZW F1 mice. The immunosuppressive properties of these classes of drugs have not been thoroughly tested, and warrant further testing before any recommendations can be made of their efficacy in lupus nephritis.
Cell-depleting and anti-cytokine therapyRituximab is a chimeric monoclonal antibody against the B-cell marker CD20 and several open clinical trials suggest that B-cell depletion with rituximab can improve clinical manifestations of lupus. However, some patients developed elevated human anti-chimeric antibodies.When a combination of rituximab, high dose steroids and cyclophosphamide was used in lupus 20 out of 21 patients experienced effective B-cell depletion, with 9 patients remaining off immunosuppressive therapy at a follow-up of 12-46 months. In another report, 9 patients with class III and IV nephritis treated with rituximab and steroids also described good efficacy, with partial remission in 80% and complete remission in 50% (30%). These encouraging results need to be confirmed by randomized controlled studies.
Oral tolerance therapyOral tolerance therapy may have a role in the treatment of lupus if the offending antigens can be identified. A phase III trial of 317 randomised patients has been completed using the B-cell tolerogen LJP394. Preliminary results indicate a sustained reduction in anti-dsDNA antibodies and improvement in health-related quality of life.Additional studies are on- going.
Immune ablation and stem cell transplantationProcedure-related mortality varies among studies between 5 and 12%. The exact role of autologous stem cell transplant using high-dose cyclophosphamide for lupus has not yet been determined and cannot be recommended.
Other newer modalitiesOther potential treatments for lupus nephritis include neutralizing antibodies to anti-C5 complement, anti-BAFF (B cell activation factor), a chemokine receptor CCR1 antagonist and human recombinant DNase. These are in the preliminary stages of development.
Prognostic factors in lupus nephritisKnowledge about prognosis assists physicians in their choice of treatment and provides patients with information on the possible outcomes. Patients with proliferative glomerulonephritis (WHO III and IV) tend to have a worse outcome for renal function when compared to patients with milder lesions. The combination of severe active and chronic histological changes on a renal biopsy is also reported to adversely affect outcome. Patients without chronic histological changes, even in the face of active lupus nephritis, had a lower risk of developing renal failure with 90% or more remaining free of renal failure after 10 years. A number of clinical variables are associated with a greater probability of renal progression in lupus nephritis. These include: black race, low haematocrit, raised serum creatinine level, presence of hypertension, high urinary protein excretion, low C3 complement and poor socioeconomic status. Failure to respond to prednisolone and cyclophosphamide are also predictors of subsequent development of renal failure, as are nephritic flares.
Dialysis and TransplantationBetween 17-30% of patients with lupus nephritis develop end stage renal failure by 10 years. Both haemodialysis and continuous ambulatory peritoneal dialysis are well tolerated and there is tendency for lupus disease activity to diminish after the start of dialysis. If there is no overt disease activity, immunosuppressants in patients on dialysis may be discontinued and a small dose of prednisolone continued. Overall survival on dialysis is good with a 75% survival at 10 years. Graft survival and function in patients with lupus after transplantation are comparable to those obtained in patients with other diseases and recurrence of lupus nephritis is uncommon after transplantation.
ConclusionAny patient with symptoms of organic dysfunction (Table 1), supported by laboratory evidence of dysfunction, must be referred to a specialist so that the diagnosis can be established, disease severity assessed and a management plan formulated. Immunosuppressive therapy and corticosteroids have, undoubtedly, had an impact on renal preservation in patients with severe lupus nephritis. No consensus has yet been reached on a number of therapeutic issues including: the optimal induction drug regimen, optimal regimen for treating relapse and maintenance therapy to prevent relapses. However, it is hoped that early diagnosis and treatment will prevent progression of renal damage, which can develop rapidly in those with severe lupus nephritis.
Complications of drug treatment account for much of the morbidity that develops in lupus, in particular, complications of high dose or chronic corticosteroids (infection, osteonecrosis, osteoporosis, coronary artery disease) and cyclophosphamide (infection, sterility, bladder toxicity and malignancy). It is encouraging that mycophenolate has a less toxic side-effect profile than cyclophosphamide. The risk of infections is also substantial in patients taking corticosteroids and immunosuppressants. Infections should be treated with appropriate antibiotics and immunosuppressants will need to be temporarily withdrawn if there is overwhelming sepsis or neutropenia.
Both the hospital physicians and general practitioners have a very important role to play in the management of these patients, especially in monitoring drug toxicity, blood pressure and renal function. Regular clinical review of the patient’s condition, together with laboratory tests and urinalysis to detect marrow depression, disease activity or progressive disease is mandatory in the management of these patients.
Table 1 - Major organ involvement in Lupus
Mesangial, focal, and diffuse proliferative glomerulonephritis
Diffuse neurologic syndromes (organic brain syndromes, psychosis, affective disorders, meningitis)
Focal neurologic syndromes (seizures, cerebrovascular events, transverse myelitis)
Movement disorders (chorea, cerebellar ataxia, Parkinson-like)
Peripheral neuropathy (symmetric sensorimotor, mononeuritis multiplex, Guillain-Barré)
Parenchymal disorders (pneumonitis, alveolar haemorrhage, bronchiolitis obliterans)
Vascular (pulmonary hypertension, pulmonary embolism)
Shrinking lung syndrome
Inflammatory bowel disease
Table 2 - The World Health Organisation Classification of Lupus Nephritis
(a) Nil (by all techniques)
(b) Normal by light microscopy but deposits by electron immunofluorescence microscopy
Mesangial glomerulonephritis (20%)
(a) Mesangial widening and/or mild hypercellularity
(b) Moderate hypercellularity
Focal proliferative glomerulonephritis (25%)
(a) "Active" necrotising lesions
(b) "Active" and sclerosing lesions
(c) Sclerosing lesions
Diffuse proliferative glomerulonephritis (37%)
(a) Without segmental lesions
(b) With "active" necrotising lesions
(c) With "active" and sclerosing lesions
(d) With sclerosing lesions
Membranous glomerulonephritis (13%)
(a) Pure membranous glomerulonephritis
(b) Associated with lesions of category II (a or b)
(c) Associated with lesions of category III (a - c)*
(d) Associated with lesions of category IV (a - d)*
Table 3. Investigations in Systemic Lupus Erythematosus
Erythrocyte sedimentation rate
Long half life
Non-specific and may not mirror disease activity
Frequently elevated during active SLE
C-reactive protein level
Raised levels are suggestive of superimposed infection
Complements, C3 and C4
Levels are decreased in active disease
anti-histone (in drug-induced SLE)
anti-Ro and anti-La (also in Sjögren's syndrome)
Tests of renal function
Urea, electrolytes and creatinine levels
24-hour urine protein level
Cardiac and pulmonary investigations, if evidence of involvement
Consultant Physician and Honorary Senior Lecturer
Department of Nephrology
Plymouth, PL6 8DH
Department of Nephrology
Queen Elizabeth Hospital
Birmingham, B15 2TH