Objectives To develop recommendations for the management of adult and paediatric lupus nephritis (LN).
Methods The available evidence was systematically reviewed using the PubMed database. A modified Delphi method was used to compile questions, elicit expert opinions and reach consensus.
Results Immunosuppressive treatment should be guided by renal biopsy, and aiming for complete renal response (proteinuria <0.5 g/24 h with normal or near-normal renal function). Hydroxychloroquine is recommended for all patients with LN. Because of a more favourable efficacy/toxicity ratio, as initial treatment for patients with class III–IVA or A/C (±V) LN according to the International Society of Nephrology/Renal Pathology Society 2003 classification, mycophenolic acid (MPA) or low-dose intravenous cyclophosphamide (CY) in combination with glucocorticoids is recommended. In patients with adverse clinical or histological features, CY can be prescribed at higher doses, while azathioprine is an alternative for milder cases. For pure class V LN with nephrotic-range proteinuria, MPA in combination with oral glucocorticoids is recommended as initial treatment. In patients improving after initial treatment, subsequent immunosuppression with MPA or azathioprine is recommended for at least 3 years; in such cases, initial treatment with MPA should be followed by MPA. For MPA or CY failures, switching to the other agent, or to rituximab, is the suggested course of action. In anticipation of pregnancy, patients should be switched to appropriate medications without reducing the intensity of treatment. There is no evidence to suggest that management of LN should differ in children versus adults.
Conclusions Recommendations for the management of LN were developed using an evidence-based approach followed by expert consensus.
This is an open-access article distributed under the terms of the Creative Commons Attribution Non-commercial License, which permits use, distribution, and reproduction in any medium, provided the original work is properly cited, the use is non commercial and is otherwise in compliance with the license. See: http://creativecommons.org/licenses/by-nc/3.0/ and http://creativecommons.org/licenses/by-nc/3.0/legalcode
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Approximately 50% of patients with systemic lupus erythematosus (SLE) will develop lupus nephritis (LN), which increases the risks for renal failure, cardiovascular disease and death. In 2008, we published the first European League Against Rheumatism (EULAR) recommendations on the management of SLE.1 Since then, several controlled trials have been published upon which updated recommendations can be based. The realisation that in the care of patients with LN internists/rheumatologists and nephrologists are involved, prompted us to develop recommendations for LN under the joint auspices of the EULAR and the European Renal Association–European Dialysis and Transplant Association (ERA-EDTA), with experts from both disciplines. The panel was enriched with renal pathologists and paediatricians with expertise on LN.
We followed the EULAR standardised operating procedures2 and the Appraisal of Guidelines Research and Evaluation instrument. We selected a list of questions by a modified Delphi method further edited for literature search, followed by a systematic search of the PubMed database (web-only appendix tables 1 and 2); all English language publications up to December 2011 were considered. We further refined retrieved items based on abstract and/or full-text content, and the number of patients (requiring n≥30 for diagnosis, monitoring, prognosis; n ≥ 10 for treatment). A detailed presentation of the literature review is provided in web-only appendix table 3. Evidence was categorised based on the design and validity of available studies and the strength of the statements was graded. After discussions, the committee arrived at 28 final statements rated individually by each member (tables 1 and 2).
Results and discussion
Indications for first renal biopsy in SLE
Because of the potentially aggressive nature of LN, the thresholds for performing a renal biopsy should be low. Any sign of renal involvement—in particular, reproducible proteinuria ≥0.5 g/24 h especially with glomerular haematuria and/or cellular casts—can be an indication for biopsy. Clinical, serological or laboratory tests cannot accurately predict histological findings. Although clinically relevant biopsy findings are more common in the presence of significant proteinuria, a biopsy may also be considered in cases of persisting isolated glomerular haematuria, isolated leucocyturia (after other causes, such as infection or drugs are excluded),3 ,4 and the rare occurrence of unexplained renal insufficiency with normal urinary findings. Lower glomerular filtration rate (GFR) is associated with chronic histological lesions and faster rate of decline in GFR.5–9 Methods for estimating GFR such as the Cockcroft–Gault and the Modification of Diet in Renal Disease equations in adults or the Schwartz formula in children, although not fully validated in SLE,10 ,11 are acceptable in clinical practice. For GFR <30 ml/min the decision for biopsy should be based on normal kidney size (>9 cm length in adults) and/or evidence of renal disease activity, in particular proteinuria and active urinary sediment (dysmorphic red blood cells (glomerular haematuria), white blood cells and/or cellular casts). Biopsy should be performed within the first month after disease onset, preferably before the institution of immunosuppressive treatment, unless contraindicated.12–14 Treatment with high-dose glucocorticoids should not be delayed if a renal biopsy cannot be readily performed.
Pathological assessment of renal biopsy
We recommend using the International Society of Nephrology/Renal Pathology Society 2003 classification system15–17 with assessment of active and chronic glomerular and tubulointerstitial changes,18–21 and of vascular lesions associated with anti-phospholipid antibodies/syndrome.22 ,23 An adequate sample of ≥8 glomeruli should be examined under light microscopy15 ,24 with haematoxylin and eosin, periodic acid-Schiff, Masson's trichrome and silver stain. Immunofluorescence or immunohistochemistry for immunoglobulin and complement deposits (IgG, IgA, IgM, C3, C1q, κ and λ light chains) is recommended.12 ,21 ,25 ,26 Electron microscopy facilitates the recognition of proliferative and membranous lesions and should be performed if possible.19 ,27–29
Indications and goals of immunosuppressive treatment in LN
Ultimate goals of treatment are long-term preservation of renal function, prevention of flares, avoidance of treatment-related harms, and improved quality of life and survival. Treatment must be based on a shared decision between patient and doctor. Immunosuppressive treatment is generally not indicated in classes I and VI LN, unless necessitated by extra-renal lupus activity.30–32
Treatment should aim for complete renal response, defined as urine protein:creatinine ratio (UPCR) <50 mg/mmol (roughly equivalent to proteinuria <0.5 g/24 h) and normal or near-normal (within 10% of normal GFR if previously abnormal) GFR. Partial renal response, defined as ≥50% reduction in proteinuria to subnephrotic levels and normal or near-normal GFR, should be achieved preferably by 6 months and no later than 12 months following treatment initiation.9 ,33–35 Improvement includes any reduction in proteinuria and normalisation or stabilisation of GFR. Although partial response carries worse prognosis than complete response,34 ,36 ,37 it may be an acceptable outcome when all treatments have been exhausted or cannot be used due to high individual risks for toxicity. Following response, patients may experience nephritic or proteinuric flares, the former having more adverse impact on renal outcomes.34 ,37–39 Nephritic flares include reproducible increase of serum creatinine by ≥30% (or, decrease in GFR by ≥10%) and active urine sediment with increase in glomerular haematuria by ≥10 red blood cells per high power field, irrespective of changes in proteinuria; proteinuric flares include reproducible doubling of UPCR to >100 mg/mmol after complete response or reproducible doubling of UPCR to >200 mg/mmol after partial response.34 ,37 ,38
Treatment of adult LN
Patients with LN should be managed, if possible, in experienced centres.40 Early trials of immunosuppressive agents have highlighted the importance of long-term (beyond 5 years) follow-up in demonstrating differences in ‘hard’ outcomes such as doubling of serum creatinine, end-stage renal disease (ESRD) and death.41–43 Such outcomes, however, are not frequent and may occur late in the course of LN. Intermediate outcome measures, such as renal response and flares, occurring in the majority of patients within the first 2 years after treatment initiation, correlate with hard outcomes in studies with long-term follow-up and are commonly used as endpoints in trials.9 ,33–35 ,37–39 ,44 Correlation does not guarantee surrogacy of these outcomes for all patients, some of whom may still have hard outcomes diverging from their intermediate outcomes.
To date, long-term data are not available for MPA (box 1). Nonetheless, the publication of the Aspreva Lupus Management Study (ALMS) trial,45 the largest trial in LN showing comparable response rates between MPA (target mycophenolate mofetil (MMF) dose 3 g/day) and intravenous cyclophosphamide (CY) (monthly pulses 0.5–1 g/m2), both administered for 6 months, together with the ease of administration and the more favourable gonadal toxicity profile of the former,46–48 formed the basis for recommending MPA as initial treatment for most cases of class III–IV LN. Evidence from transplantation medicine49 ,50 and a single randomised controlled trial (RCT) in LN51 suggests that MMF and enteric-coated mycophenolic acid sodium (eMPA) are likely to be equally efficacious. To this end, and while awaiting further validation, the Committee felt that either MPA formulation can be used in treatment of LN, with 720 mg dose eMPA roughly equivalent to 1 g dose of MMF. We also recommend low-dose intravenous CY (total dose 3 g over 3 months) in combination with glucocorticoids (0.5 mg/kg/day) as initial treatment of class III–IV (±V) LN in Caucasians based on better efficacy/toxicity ratio than high-dose intravenous CY.44 ,52
Special training sessions for renal pathologists to improve the interpretation of renal biopsy findings in lupus nephritis (LN) and enhance interobserver agreement
Development and validation of biomarkers which will better reflect kidney biopsy findings and renal disease activity and severity
Long-term (beyond 5 years) efficacy and safety data for mycophenolic acid
Provide data to guide duration of immunosuppressive treatment beyond 3 years
Define the role of adding calcineurin inhibitors, rituximab or belimumab to standard immunosuppressive treatment in cases with residual renal disease
Need for more data on switching regimens in cases of treatment failure
Larger studies with extended follow-up are needed to assess the prognostic significance of anti-phospholipid syndrome (APS)-associated nephropathy (APSN) and coexistence of anti-phospholipid antibodies in LN
Need for controlled trials to assess the role of antiplatelet/anticoagulant regimes in APSN
Need for randomised controlled trials (RCTs) in paediatric LN and the need to have very long follow-up (beyond 10–15 years) to fully assess the impact of the various treatment strategies and modalities in children
A single RCT in patients with pure class V LN demonstrated that the combination of glucocorticoids with intravenous CY (6 bimonthly pulses 0.5–1 g/m2) was more efficacious than glucocorticoids alone; the combination of glucocorticoids with ciclosporin was also efficacious but was associated with significantly more relapses of nephrotic syndrome than CY.53 Moreover, combined analysis of two other RCTs in the subgroup of patients with pure class V LN showed a comparable antiproteinuric effect of MPA versus high-dose intravenous CY.54 By extrapolation from these studies, and based on the more favourable gonadal toxicity profile of MPA compared to CY, we recommend MPA as initial treatment for most cases of class V LN and nephrotic-range proteinuria. The low-dose CY regimen has not been tested in pure class V LN.
Subgroup analysis suggests that MPA may have greater efficacy in patients of African descent;45 ,55 further confirmation is needed before issuing a recommendation favouring MPA in these patients. Post hoc analysis in 32 patients in ALMS with baseline GFR<30 ml/min/1.73 m2,45 and evidence from 2 controlled studies in severe histological forms of LN,56 ,57 support the use of MPA in patients with impaired renal function or crescents. Only high-dose intravenous CY has demonstrated efficacy in a RCT specifically designed to include severe nephritic cases with GFR 25–80 ml/min or with crescents/necrosis in >25% of glomeruli.58 Data from a RCT59 and the 10-year follow-up60 suggest that azathioprine can be used in class III–IV LN albeit at an increased risk for renal relapse (HR 4.5), thus the committee recommends it for milder cases (preserved renal function and no adverse histological findings).
Intravenous methylprednisolone (MP) pulses are recommended as part of the initial treatment regimen by extrapolation from controlled studies,43 ,52 ,61 ,62 to decrease cumulative glucocorticoid dose and associated harms. Higher initial glucocorticoid dose (oral prednisone 0.7–1 mg/kg/day) may be used in severe renal or extra-renal lupus, or when intravenous MP treatment is not feasible. Clinical experience suggests that a further course of three intravenous MP pulses can be considered in patients failing to improve within the first 3 months.
For class II LN with proteinuria >1 g/24 h despite renin-angiotensin-aldosterone system (RAAS) blockade, especially in the presence of glomerular haematuria, we recommend low-to-moderate doses of glucocorticoids (prednisone 0.25–0.5 mg/kg/day) alone or in combination with azathioprine (1–2 mg/kg/day), if needed, as steroid-sparing agent. Glucocorticoids alone or in combination with immunosuppressive agents may also be considered in cases of class I LN with podocytopathy on the electron microscopy (minimal change disease)63 ,64 or interstitial nephritis.65 ,66
For patients improving after initial treatment, we recommend subsequent immunosuppression to consolidate renal response and prevent flares. Although among patients from European ancestries azathioprine and MPA were equivalent after initial treatment with low-dose intravenous CY,67 a larger RCT suggested a difference between the two drugs in favour of MPA after initial response to either MPA or intravenous CY (monthly pulses 0.5–1 g/m2).68 In this trial, sequential use of azathioprine after MPA resulted in more treatment failures as compared to MPA followed by MPA. The committee therefore recommends continuation of MPA if the drug was successful as initial treatment. Calcineurin inhibitors can be considered in selected cases with preserved renal function based on evidence from RCTs.69–71 Intravenous CY, pulsed every 3 months, may be used in selected cases43 ,58 ,72 but exposure to CY should be minimised, especially in women at risk for amenorrhoea and infertility73 or men planning to father children.
There is no data to guide duration of treatment beyond 3 years;67 ,68 continuing treatment for longer time periods should be individualised with an effort first to withdraw glucocorticoids before immunosuppressive agents. Gradual drug dosage titration may be attempted to ensure the best possible efficacy/toxicity ratio. MPA dose often needs titration to reduce toxicity (doses 1–2 g/day can be effective for long-term treatment). Monitoring MPA blood levels to minimise harm and increase efficacy is under investigation74–76 but it should be considered in cases with GFR <30 ml/min.
Complete renal response can take up to 2 years to reach with <30% to 40% of patients achieving this outcome within the first 6 months of treatment.48 ,59 Switching to an alternative agent is recommended for patients who fail to improve within 3–4 months, or do not achieve partial response after 6–12 months, or complete response after 2 years of treatment. For patients not responding to MPA or CY, evidence from uncontrolled studies suggests that treatment may be switched from MPA to CY, from CY to MPA,77 ,78 or that rituximab (anti-CD20 mAb) may be given either as add-on treatment or as monotherapy.79 ,80 Additional options include calcineurin inhibitors (ciclosporin A, tacrolimus),81–83 intravenous immunoglobulin,84 plasma exchange for rapidly progressive glomerulonephritis,49 ,85 or immunoadsorption for patients who have failed or cannot tolerate other treatments.86 ,87 Data on leflunomide are limited.88
Adjunctive treatment in patients with LN
We recommend control of cardiovascular disease risk factors in a manner similar to patients who do not have SLE with chronic kidney disease, although benefit has not been demonstrated specifically in SLE.89 Complications of chronic renal insufficiency (anaemia, cardiovascular disease, metabolic bone disease) should also be managed as in patients who do not have SLE. RAAS blockers are recommended as preferred treatment in all patients who are not pregnant with significant proteinuria or hypertension, based on: (a) evidence for their antihypertensive, antiproteinuric and renoprotective effect,90–92 and, (b) lack of data on the comparative efficacy of other classes of antihypertensive agents in LN. Their dose is titrated for maximum antiproteinuric effect while monitoring blood pressure (target level <130/80 mm Hg), serum potassium and GFR levels. Epidemiological studies93 ,94 and the follow-up of a controlled trial95 demonstrate that hydroxychloroquine use is associated with higher rates of renal response, fewer renal relapses and reduced accrual of renal damage. Hydroxychloroquine (6.5 mg/kg/day or 400 mg/day, whichever is lower) is generally safe in patients with normal baseline ophthalmological examination; dose adjustments may be necessary in patients with GFR <30 ml/min. Annual ophthalmological screening begins after 5 years of treatment or sooner if there are risk factors for retinal damage.96 Patients should also be immunised with non-live vaccines according to the EULAR recommendations.97 ,98
Monitoring and prognosis of LN
Patients should be monitored regularly according to EULAR recommendations,99 including annual examination of cervicovaginal smear in women100 ,101 and measurement of serum immunoglobulins at baseline and then annually in patients who receive immunosuppressive treatment to assess risk of infection. Monitoring of body weight, blood pressure, serum creatinine and estimated GFR, serum albumin, proteinuria, urinary sediment (microscopic evaluation), serum C3/C4, serum anti-dsDNA antibody levels and complete blood cell count are used to define activity and evaluate response to treatment although their individual predictive value for hard outcomes at particular time points is modest.
Spot UPCR measured on first morning void urine sample is a valid and conveniently repeatable measure for measuring proteinuria in children and monitoring within-patient changes in adults.102–104 Timed (12 h or 24 h) urine collections may also be considered at baseline and when major therapeutic changes are considered. Reappearance of urine cellular casts has >80% sensitivity and specificity for renal flares.105
Although serum C3 has generally higher sensitivity than serum C4 (72% to 85% vs 28% to 74%), both tests have modest specificity for active LN.106 ,107 The diagnostic accuracy of serum anti-dsDNA is also modest with positive and negative likelihood ratios ranging from 1.5–4.8 and 0.3–0.8, respectively. Farr and ELISA methods are both acceptable, although the former yields higher sensitivity and specificity rates.106 ,108–110 Anti-C1q106 ,111 and anti-nucleosome112–114 antibodies have higher sensitivity and specificity for active nephritis but further standardisation and validation are required. Changes in serological tests are more important predictors of concurrent or impending LN flare than their absolute levels but should be repeated no more than monthly. In the absence of proteinuria, active serology (decreasing C3/C4 and/or increasing anti-dsDNA) and/or urine sediment is not an indication for pre-emptive treatment but dictates closer monitoring of patients. Repeat renal biopsy provides additional prognostic information115–118 and can assist therapeutic decisions in patients with relapse of nephritis after complete renal response, or with refractory disease. It can also be used in the context of a clinical trial to monitor treatment efficacy and changes in chronicity scores.8 ,119
Management of ESRD in LN
Despite immunosuppressive treatment, 10% to 30% of patients with LN will progress to ESRD within 15 years of diagnosis. Infections (including peritonitis) may occur in patients with active disease still on immunosuppressive treatment, and contribute to morbidity and mortality.120–123 Although clinical and serological activity tend to subside in most patients with ESRD on dialysis,120 ,124–126 flares of renal or extra-renal lupus can occur.127–130
Comparative studies131 ,132 and cases series133 ,134 support that patients with SLE are good candidates for renal transplantation performed when clinical (and ideally, serological) lupus activity is absent, or at a low level, for at least 3–6 months135; best results are obtained with living donor136–138 and pre-emptive transplantation.139 Patients with moderate to high titres of anti-phospholipid antibodies are at increased risk for thrombotic complications and may receive anticoagulants perioperatively.140–143 Post-transplantation recurrent LN, although difficult to treat, is a rare cause of renal allograft loss.136 ,144 ,145
Anti-phospholipid syndrome (APS)-associated nephropathy (APSN) in SLE
Anti-phospholipid antibodies (anti-cardiolipin antibodies, anti-β2-glycoprotein I antibodies, lupus anticoagulant) may be associated with a distinct type of vascular nephropathy (APSN) with adverse prognostic factors such as hypertension, impaired renal function and interstitial fibrosis.146–149 Histological lesions of APSN are present in 20% to 30% of patients with SLE146 ,150 and include thrombotic microangiopathy and chronic lesions such as fibrous intimal hyperplasia, organising thrombi with recanalisation, focal cortical atrophy and fibrous occlusions of arteries/arterioles, thus, need to be distinguished from thrombotic thrombocytopenic purpura/haemolytic uraemic syndrome and malignant hypertension. In spite of lack of evidence from controlled studies, hydroxychloroquine and/or antiplatelet/anticoagulant treatment can be considered in combination with immunosuppressive treatment if nephritis is present. Patients with definite APS should receive anticoagulation treatment.151
LN and pregnancy
Pregnancy may be planned in patients with inactive lupus and UPCR <50 mg/mmol for the preceding 6 months, with GFR that should preferably be >50 ml/min. Patients with LN who are pregnant should ideally be followed by a multidisciplinary team. Stable renal disease is treated with the same drugs that are recommended as acceptable during prepregnancy counselling (hydroxychloroquine, prednisone, azathioprine). Hydroxychloroquine should be continued152 ,153 or even instituted if immunosuppressive agents need to be stopped. MPA or CY should not be used in the last 3 months, and biological agents for at least 4 months—dependent upon the agent used before conception. Blood pressure should be controlled without RAAS blockers at the time of conception if possible, due to their potential teratogenic effect during the first trimester, or with switching to other agents such as nifedipine or labetalol as soon as pregnancy is confirmed.154 ,155 Acetyl-salicylic acid is recommended to reduce the risk for pre-eclampsia.156 Patients with APS are at increased risk for adverse pregnancy outcomes154 ,157 ,158 and should be considered for anticoagulation with low-molecular-weight heparin and/or acetyl-salicylic acid depending on their history of obstetric and/or thrombotic events.151 Warfarin must be discontinued as soon as pregnancy is confirmed. Patients with nephrotic-range proteinuria are also candidates for anticoagulation.
For monitoring, any fall in serum C3/C4 is significant given than levels usually rise during pregnancy;159 additional investigation may be needed to rule out pre-eclampsia before diagnosing exacerbation of renal disease.160 For active disease or pre-eclampsia, combined care with obstetricians is recommended.158 Close surveillance for renal flare post partum is essential. In addition to acceptable medications used in stable LN, refractory cases can also be treated with calcineurin inhibitors, intravenous immunoglobulin, immunoadsorption and possibly plasma exchange, according to disease severity.156 ,161
Management of paediatric LN
Children are at increased risk for renal involvement compared to adults with SLE (OR 1.5–2.4), and nephritis often is a presenting feature of paediatric SLE. Together with elevated blood pressure, fever, lymphadenopathy, skin and joint manifestations,162 children with LN tend to have more active disease over time, receive more intensive immunosuppressive treatment and accrue more damage, often related to glucocorticoid toxicity, compared to adults.163–168 The diagnosis, management and monitoring is based on extrapolation from evidence in adults, and on the limited, non-randomised, evidence in children with LN.169–172 Additional considerations include the negative effect of disease activity and glucocorticoids on linear growth, and the modification of body image induced by treatment. This may represent major psychological burden especially in adolescents building their self-esteem and affecting treatment compliance.
Support for this work was provided via grants from the EULAR Standing Committee on Clinical Affairs (ESCCA) and the ERA-EDTA.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online appendix 1
- Data supplement 1 - Online appendix 2
- Data supplement 1 - Online appendix 3
Contributors GKB performed the systematic review of the literature, organised the results and their presentation and also drafted the manuscript. JPAI is the clinical epidemiologist who supervised the analysis of the literature findings and the grading of evidence. DTB and DJ supervised the project and chaired discussions. All coauthors contributed to discussions, drafted the statements and reviewed the manuscript.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.