Appendix Table 3.Questions for the literature search and results from the systematic review of the literature
Q1. What are the indications and optimal timing for first renal biopsy in SLE patients and are they different compared to non-SLE patients?
Item / Example(s) of outcome or effect size / No. studies / Design1 / ReferencesAbnormal SCr (≥130 μmol/L [1.4 mg/dL]) and/or GFR levels (<80 ml/min/1.73m2) correlate with adverse histological forms of LN / SCr ≥130 μmol/L associated with OR 2.1 for class IV LN / 24 / III / 1-24
Abnormal urinalysis (proteinuria >0.5 g/24-hr and/or haematuria and/or cellular casts) with normal SCris associated with class III-IV LN / Proteinuria >0.5 g/24-hr or haematuria or cellular casts: 70% class III-IV LN at biopsy; isolated haematuria (>5 RBC/hpf) with proteinuria <0.5 g/24-hr: 22% active class III-IV LN at biopsy / 26 / III / 1,3,6,7,9,11,13-20,23,25-35
Clinically ‘silent’ LN (proteinuria <0.3-0.5 g/24-hr, inactive urinary sediment, normal GFR) is associated with class I-II LN and a favorable outcome / Prevalence of histologic class III-V LN: 86% in ‘overt’ versus 36% in ‘silent’ cases (defined as proteinuria <0.3 g/24-hr with inactive urinary sediment andGFR >70 ml/min//1.73m2) / 10 / III / 36-45
Delay in kidney biopsy and diagnosis of LN is associated with progression of renal disease and histological lesions, reduced rates of renal remission after treatment, and increased risk for ESRD / Delay >6 months is associated with adjusted HR 4.2–9.3 for ESRD / 9 / II / 3,13,28,46-51
Q2. Evaluation of the ISN/RPS 2003 classification system. What is the prognostic significance of renal biopsy findings?
Item / Example(s) of outcome or effect size / No. studies / Design / ReferencesImproved inter-observer reproducibility with the ISN/RPS 2003 classification / Kappa coefficient for inter-observer reproducibility is 0.53 versus 0.44 with the WHO classification; cases re-classified from WHO class III into ISN/RPS class IV had lower rates of response to therapy and increased risk for progression to ESRD / 6 / III / 5,16,52-55
The histological class of nephritis (defined by the WHO or the ISN/RPS classification) correlates with patient and renal outcome (death, ESRD, doubling SCr, development of chronic renal insufficiency) / Proliferative (class III-IV) versus non-proliferative LN is associated with RR 3.0 for death; class IV versus class III LN is associated with HR 4.3 for ESRD; class V LN has worse outcome than class I-II (abnormal renal function in 15% and ESRD in 5% during long-term follow-up); class V with superimposed proliferative lesions (class V+III-IV) has worse outcome than pure class V (ESRD rate 36% versus 12%) / 52 / II / 1,3-6,11,14,16,28,47,48,56-96
ISN/RPS class IV-G versus IV-S is associated with more severe clinical disease and/or worse histological lesions / Increased endo-capillary hyper-cellularity, interstitial inflammation, activity lesions; no definitive difference in renal outcome / 11 / III / 5,12,17,18,33,55,97-101
Prognostic significance of kidney deposits detected by immunofluorescence or electron microscopy / Intensity/extension of deposits correlates with clinical & histological severity of renal disease, and renal outcome; intra-membranous electron-dense deposits are associated (RR 4.8) with death / 19 / II / 1,7,46,47,61,94,102-114
Prognostic significance of activity lesions on kidney biopsy (hypercellularity, leukocyte exudation, karyorrhexis/fibrinoid necrosis, cellular crescents, hyaline deposits, interstitial inflammation) / The composite AI correlates with poor renal outcome (increased relapse rates, doubling SCr, ESRD, death; HR 1.4 per 1-unit); cellular crescents >30% of glomeruli is associated with RR 7.2 for doubling SCr; ≥25% interstitial infiltration is associated with HR 1.9 for doubling SCr or ESRD / 44 / II / 1,7-9,11,13,15,25,46-48,56-58,65,79,101-103,106,113,115-12516,28,91,92,100,126-128, 50,94,110,111
Prognostic significance of chronicity lesions on kidney biopsy (glomerular sclerosis, fibrous crescents, tubular atrophy, interstitial fibrosis) / The composite CI correlates with poor response to immunosuppressive therapy, and poor long-term renal outcome (increased relapse rates, doubling SCr, ESRD, death; HR 1.18 per 1-unit for doubling Scr or death); glomerular sclerosis ≥25% glomeruli is associated (RR 3.3) with ESRD; interstitial fibrosis is associated (HR 2.0) with doubling SCr or ESRD; tubular atrophy ≥25% is associated (HR 2.4) with doubling SCr or ESRD / 60 / II / 1,3,8,10,11,13,17,23,25,28,47,48,56,57,61,63,65,73,75,76,79,81,84,87,90,92,100-103,110,111,113,115,119-127,129-145
Prognostic significance of tubulointerstitial lesions on kidney biopsy / The composite tubulo-interstitial index (interstitial infiltrates + tubulorrhexis + interstitial sclerosis) is associated with higher proteinuria and SCr levels, and increased risk for ESRD (HR 72 in patients with class V LN) / 15 / II / 7-9,15,46,47,58,79,101,102,106,115-118
Prognostic significance of renal vascular lesions (lupus vasculopathy, HUS/TTP/malignant hypertension-like lesions, vasculitis, arteriosclerosis) / Association with higher SCr levels, increased rates of hypertension, lower 5- and 10-yr kidney survival rates (74% and 58% vs. 90% and 86% in patients without vascular lesions) / 3 / III / 67,146,147
Additional prognostic value of renal histology to renal clinical disease severity / The proportion of sclerosed glomeruli combined with SCr levels at 2 years post initiation of therapy may predict 59% of the variation in SCr levels at last follow-up in proliferative LN / 4 / II / 23,46,47,70
Q3. How well do biochemistry tests, serological tests, urinalysis, biomarkers correlate with renal biopsy findings in lupus nephritis?
Item / Example(s) of outcome or effect size / No. studies / Design / ReferencesDiagnostic value of SCr and/or GFR (estimated or measured) levels / Higher SCr and/or lower GFR levels correlate with more advanced histological class (class IV>III>other; IV-G>IV-S), higher biopsy AI and CI (r = 0.27-0.46), tubulo-interstitial lesions, kidney deposits, and presence of cellular crescents; in patientswithSCr ≤1.4 mg/dL, those with MDRD-GFR below 60 ml/min/1.73m2 had more chronic lesions in renal biopsy / 26 / III / 1-24,30,103
Diagnostic value of hemoglobin levels / Lower hemoglobin levels correlate with more advanced histological class (Hb <9 g/dLis associated with OR 3.2 for class IV LN) / 6 / III / 6,7,14,17,18,33
Diagnostic value of serum complement levels / Lower levels at the time of diagnosis correlate modestly with more advanced histology (proliferative < non-proliferative LN; class IV < III; IV-G < IV-S); correlation with karryorrhexis, endocapillary proliferation, and CI (r = -0.32 to -0.39); there is a single study reporting better predictive ability of serum C3 versus C4 / 13 / III / 2,4,14,18,23,30,148-154
Diagnostic value of serum anti-dsDNA antibodies / Increased titers (ELISA, Crithidia, or Farr assay) at the time of diagnosis is associated with proliferative versus non-proliferative LN (OR 6.0), class IV > other; correlation with active / severe renal histology (biopsy AI, extra-capillary proliferation, karryorhexis, hyaline deposits, fibrinoid necrosis, electron-dense deposits; rcoefficient ranging 0.35 to 0.81) / 18 / III / 1,2,7,23,103,148-150,152-161
Diagnostic value of serum anti-C1q antibodies / Increased titers (ELISA) are associated with active proliferative (class III-IV) LN (PPV 68%, NPV 98%); higher titers in class IV > other classes); correlation with biopsy AI and CI, glomerular infiltration, endo-capillary proliferation, fibrinoid necrosis, crescents (rcoefficient ranging 0.32 to 0.56) / 10 / III / 18,63,149,155,159,162-166
Diagnostic value of other serological tests (anti-chromatin, anti-histone, anti-nucleosome, anti-ribosomal P, anti-RNP, anti-Sm, ANCA, anti-endothelial cell antibodies) / Limited diagnostic value; higher titers of anti-endothelial cell IgG in class IV vs. other classes of LN / 19 / III / 17,18,149,151,156,159,160,167-178
Diagnostic value of proteinuria (24-hr urine collection or random spot urine protein to creatinineratio) / Higher proteinuria levels correlate with more advanced histological lesions (class IV-VIIIII; IV-GIV-S), higher biopsy AI and CI, increased glomerular and tubulo-interstitial lesions; proteinuria >0.2 g/24-hr is associated (OR 4.4) with class IV LN; nephrotic-range proteinuria is common in class IV or V (± III-IV) LN / 14 / III / 1,3,6,9,11,13-16,18,19,32-34
Diagnostic value of active urine sediment (haematuria >5 RBC/hpf and/or pyuria >5 WBC/hpf and/or cellular casts); see also Q1 / More common in active class III-IV LN (OR 6.0), class IV-G versus IV-S, crescentic versus non-crescentic forms of LN; correlation with proteinuria levels and higherbiopsy AI and CI; isolated abnormalities (in absence of proteinuria >0.5 g/24-hr) are less predictive for active class III-IV LN / 11 / III / 1,7,9,11,15,17,19,23,29,33,35
Diagnostic value of serum biomarkers (ICAM-1, VCAM-1, fribrinolytic split products, IL-18, nitrate, IL-12p70, IFN-γ, IL-2-receptor, IFN-γ+/ΙL-4+ PBMCs [flow cytometry]) / Limited diagnostic value / 9 / III / 149,179-186
Diagnostic value of urine biomarkers (cytokines [IL-6, IL-12, IL-10, IL-2], TGFβ, ICAM-1, VCAM-1, T-bet, GATA-3, RANTES, MCP-1, proteomics) / MCP-1 levels correlate with active class III-IV LN, biopsy AI and CI (n=4 studies); proteomic analysis demonstrates sensitivity 86-100% and specificity 50-100% for different histological classes of LN, and correlates with biopsy AI (r = 0.77) and CI (r = 0.87) / 12 / III / 30,182,185,187-195
Q4. What are the indications for initiating immunosuppressive therapy in lupus nephritis?
Item / Outcome and example(s) of effect size / No. studies / Design / ReferencesClass II LN / Treatment with low-to-moderate doses of GC (0.25-0.5 mg/kg/day) if proteinuria exceeds 1 g/24-hr especially in presence of glomerular haematuria; immunosuppressive therapy (usually AZA; MPA and CY have also been used) is added in cases of inadequate response to GC (± HCQ), or when unable to taper their dose / 6 / V / 59,196-200
Class III-IV LN / Combination of immunosuppressive agents with GCare more effective than GC alone in prevention of ESRD or death; effect more pronounced in class IV versus III LN / 4 / I / 201-204
13 / II / 57,139,205-215
9 / III/IV / 63,82,83,123,134,143,216,217
Class V LN / Combination of immunosuppressive agents with GC are indicated in cases with persistent proteinuria >1 g/24-hr despite optimal use of renin-angiotensin-aldosterone blockade; the use of adjunctive immunosuppression is independent predictor for remission of proteinuria in cases with nephrotic-range proteinuria / 1 / II / 218
2 / III / 219,220
8 / V / 221-228
Q5. What is the comparative evidence for the benefits and harms of therapeutic agents or interventions in patients with lupus nephritis and according to different histologic classes of kidney disease?
Item / Outcome and example(s) of effect size / No. studies / Design / ReferencesClass II LN / Low-to-moderate doses of GC (0.25-0.5 mg/kg/day); immunosuppressive therapy can be added (usually AZA; also MPA, CY) if proteinuria exceeds 1 g/24-hr, especially in presence of glomerular haematuria / 6 / V / 59,196-200
Class III-IV LN: initial treatment / Glucocorticoids
- GC-only regimens are less efficacious than the combination of GC and immunosuppressive agents in prevention of ESRD and/or death
14 / II / 57,139,205-215,229
9 / III/IV / 63,82,83,123,134,143,216,217
- Regimens used: high-dose GC (1 mg/kg/day ×4-6 wks, then tapered); or, pulses IV-MP (0.5-1 g ×3 days), followed by oral prednisone (0.5-0.7 mg/kg/day); both regimens have been used in RCTs
- Benefit from the use of IV-MP pulses: may expedite the reduction of abnormal SCr and/or proteinuria levels; has been associated with improved renal outcome (increased remission rates, reduced relapse rates; extrapolation from two RCTs); has been used in severe crescentic LN
1 / III / 230
3 / IV / 16,138,231
8 / V / 17,232-238
Cyclophosphamide
- Combination of CY (IV or oral) with GC is efficacious in stabilizing renal function (RR 0.63 for ESRD; RR 0.59 for doubling SCr); superiority over GC-only regimen becomesevident after 5 years of follow-up; long-term (≥10 years) follow-up data available
14 / II / 57,139,206-212,214,215,229
9 / III/IV / 63,82,83,123,134,143,216,217
11 / V / 87,119,140,198,199,239-244
- Benefit demonstrated in patients with high-risk features (high biopsy CI, impaired renal function, nephrotic syndrome, class V+IV LN, crescentic LN, rapidly progressive glomerulonephritis)
3 / III / 237,246,247
10 / V / 17,134,237,238,248-253
- Comparable efficacy of IV-CY versus oral CY; in the long-term (7 years) follow-up of a RCT comparing immunosuppressive agents versus GC alone, only the IV-CY regimen was statistically superior in individual group comparisons; in the meta-analyses of early RCTs, results were more robust for the IV-CY regimen; a single RCT comparing intermittent IV-CY versus oral CY followed by AZA found no difference in efficacy or adverse events (trial prematurely terminated); trend for more adverse reactions (herpes zoster infections, leukopenia) with oral CY; oral CY has demonstrated efficacy in class IV LN with rapidly progressive renal failure
5 / II / 57,209-211,254
2 / III / 246,247
2 / IV / 134,255
- Low-dose versus high-dose IV-CY. In White patients with moderate-risk features (class IV 70%; mean proteinuria 2.9-3.2 g/24-hr; mean biopsy CI 0.8-0.9), initial treatment with low-dose IV-CY (3g) has similar efficacy (prevention of ESRD; follow-up 10 years) and better toxicity profile (reduced infections) than high-dose IV-CY (NIH regimen; both regimens followed by AZA); in patients with high-risk features (class IV 85-90%; 35-50% Black; nephritic with GFR <80 ml/min or GFR >80 ml/min with crescents >25%; mean proteinuria 5.0 g/24-hr; mean biopsy CI 3.7-4.4), extended course (30 months) of IV-CY was associated with lower rates of doubling SCr and of renal relapse compared to short course (6 months) IV-CY
6 / II / 245,257-261
2 / IV / 134,216
6 / V / 240,250,255,262-264
- Toxicity of CY-based regimens: amenorrhea (dose- and age-dependent risk: in women aged ≥32 years D50has been calculated to 8 g/m2), infections (particularly herpes zoster), leukopenia
Azathioprine
- Early trials showed that the combination of AZA with GC is more efficacious than GC alone in preventing death (RR 0.6) but not ESRD
- In patients with class III-IV (±V) LN and moderate-risk features (70% Whites; mean SCr 109 μmol/L, mean proteinuria 3.2 g/24-hr, mean biopsy CI 2.3), the AZA/IV-MP/GC regimen had comparable rates of renal response with the IV-CY/GC regimen during the first 2 years (after a median follow-up of 6.3 years more than 80% of AZA-treated patients remained free of CY; after a median follow-up of 9.6 years there were no significant differences between the two regimens in terms of doubling SCr, ESRD, or mortality, although AZA-treated patients had HR 4.5 for renal relapse)
- Weak evidence-base for efficacy of AZA (prevention of ESRD) in severe forms of class III-IV LN
14 / II / 23,139,205,206,209,210,213,230,268-273
3 / III / 274-276
9 / V / 50,119,197,277-282
Mycophenolic acid (mycophenolatemofetil, mycophenolate sodium)
- Comparable efficacy with CY regimens in terms of induction of renal remission (rates 56-81% in MPA, 52-76% in CY), prevention of ESRD and death; characteristics of patients included in the RCTs: most (53-67%) had class IV LN, relatively preserved renal function (mean SCr 1.1-1.3 mg/dL), average proteinuria 4.1-6.2 g/24-hr (except for two trials: 1.8-2.0 g/24-hr), chronic lesions in kidney biopsies (mean biopsy CI 2.8-3.9), heterogeneous ethnic background; MMF dose: 2-3 g/day
- A single RCT demonstrating superiority of MPAover CY included more Black patients, more patients with pure class V LN, and only 62% of patients in the CY arm received all 6 pulses of IV-CY
- Significant variation in renal response rates according to race (lower response rates of IV-CY in Black/Hispanics) and region (better efficacy of MPA in regions outside Asia)
- More favorable toxicity profile of MPA versus CY (reduced amenorrhea, leukopenia, alopecia rates; trend for lower rates of infections; increased gastrointestinal events [diarrhea] with MPA versus CY)
- Open-label studies have also demonstrated efficacy and histological improvement with MPA in class III-IV LN
7 / II / 290-296
1 / III / 297
8 / V / 49,88,199,298-302
- Efficacy of enteric-coated mycophenolate sodium (1440-2160 mg/day) in class III-IV LN
1 / III / 304
2 / V / 305,306
- Crescentic LN with impaired renal function: comparable efficacy of MPA versus CY (higher response rates but also more relapses with MPA)
1 / III / 308
1 / V / 237
Ciclosporin A
- A single RCT comparing oral CsA versus IV-CY (both in combination with GC) in patients with moderate-risk features (class IV 53-67%; SCr <140 μmol/L; mean proteinuria 2.5-3.8 g/24-hr; mean biopsy CI 3.5-4.0) found comparable rates of renal response (at 9 months) and improvements in renal disease parameters (reduction in proteinuria)
- Toxicity: hypertension (very common); hirsuitism, dose adjustments may be needed due to SCr elevations
3 / V / 310-312
Tacrolimus
- Comparable short-term efficacy of tacrolimus versus pulses IV-CY (both in combination with GC) in class III-IV LN with moderate-risk features (class IV 72%; normal SCr levels; mean proteinuria 2.2-3.0 g/24-hr; mean biopsy CI 1.0-1.2).
- In patients with mixed class V+IV LN (normal SCr levels; mean proteinuria 4.1-4.4 g/24-hr, mean biopsy CI 1.3-1.4; 65% had been previously treated with MPA or CY), combination therapy with tacrolimus/MMF/prednisone was associated with higher rates of complete response at 6 months compared to pulse IV-CY therapy (50% versus 5%)
6 / V / 315-320
Plasma exchange
- No benefit of adding plasma exchange cycles to immunosuppressive therapy (AZA, IV-CY) in terms of renal remission or death rates (evidence from RCTs that included patients with class IV 40-100%; mean proteinuria 5.5-8.6 g/24-hr, impaired renal function)
- Plasma exchange / immunoadsorptioncycles in combination with GCand/or cytotoxic therapy have been used in cases with rapidly progressive glomerulonephritis or crescentic class IV LN
3 / II / 321-324
1 / III / 325
2 / IV / 326,327
2 / V / 328,329
Leflunomide
- Single non-randomized study in mild class III-IV LN has shown comparable efficacy with low-dose pulse IV-CY; increased biopsy CI at follow-up renal biopsy
1 / III / 331
1 / V / 332
Chlorambucil
- Oral chlorambucil in combination with GChas shown comparable efficacy with the combination of oral CY with GC in prevention of renal or patient death (active class III-IV ± V cases; histological improvements at follow-up renal biopsy)
1 / V / 333
Rituximab
- Combination therapy with rituximab/MMF/GC was no superior than MMF/GC in terms of renal response rates (57% versus 46% at 1 year); n=8 placebo patients versus no rituximab-treated patients required CY rescue therapy (cases with moderate-risk features: White 31%; class IV 66%; mean SCr1.0 mg/dL; mean proteinuria 4.0 g/24-hr)
- Efficacious in small case-series of active class III-IV LN (used as first-line treatment)
Belimumab
- Post-hoc analysis of RCT: reduced rates of renal flares in the high-dose drug arm versus placebo
Class III-IV LN: subsequent treatment /
- In class III-IV LN patients responding to initial immunosuppressive therapy (‘induction therapy’), absence or shorter duration of subsequent immunosuppressive therapy (‘maintenance therapy’) is associated with increased risk for renal relapse
Cyclophosphamide
- Lower dose and/or less frequently administered intermittent IV-CY (quarterly pulses) or oral CY, sequentially administered after remission induction for another 12-24 months is associated with lower rates of renal relapse, doubling SCr or ESRD; experience in patients with high-risk features (nephrotic syndrome and/or impaired renal function) with remission rates up to 72%
2 / III / 246,297
2 / IV / 134,216
10 / V / 87,136,198,239,240,243,252,263,264,341
Azathioprine
- Used for maintenance of renal remission following successful initial treatment with CY (evidence from RCTs and uncontrolled case series); also, sequentially after initial treatment with AZA (in milder LN cases)
- A single retrospective controlled study in class IV LN (mean proteinuria 5.9-6.8 g/24-hr, mean SCr 122-135 μmol/L) comparing pulse IV-CY versus oral CY followed by AZA found no difference in renal relapse or doubling SCr rates
2 / III / 247,274
3 / IV / 133,134,217
13 / V / 50,140,197,242,244,249,250,263,264,278,280,281,305
Mycophenolic acid
- Used after initial treatment with CY or MPA (sequential therapy)
- RCT comparing quarterly pulses IV-CY (0.5-1 g/m2) versus MMF (0.5-3 g/day) versus AZA (1-3 mg/kg/day) after initial treatment with pulse IV-CY in patients with high-risk features (Hispanics 50%; class IV 68-85%; mean SCr 1.5-1.7 mg/dL; mean proteinuria 4.7-5.7 g/24-hr; mean biopsy CI 1.9-3.8): during follow-up (mean 5.5 years) there were reduced renal and patient survival rates and more relapses in the IV-CY group (40% versus 32% in AZA, 15% in MMF)
- ALMS trial (42% Whites, class IV ± V 72%, composite endpoint, responders after initial treatment with either IV-CY or MMF were re-randomized to study arms): reduced renal flares with MMF versus AZA (13% versus 23% during 3-year period); MAINTAIN trial (79% Whites, class IV 59%, time to flare used as endpoint, randomization upon enrollment regardless of response to initial low-dose IV-CY regimen): comparable rates of renal flares with MMF versus AZA (19% versus 25% during 3-year period)
4 / II / 140,343-345
2 / IV / 242,346
12 / V / 49,88,199,241,251,264,299-301,305,347
Ciclosporin A
- Used after initial treatment with CY or CsA (sequential therapy); efficacious in class III-IV LN with preserved renal function and moderate-risk features
1 / V / 312
Tacrolimus
- Limited experience with prolonged (sequential) treatment of class III-IV LN
Intravenous Ig (IVIG)
- Has been used after good response to initial treatment with IV-CY, with comparable efficacy with extended pulse IV-CY treatment (LN cases with moderate-risk features: White patients, class IV 20-23%; single study)
Plasma exchange / immunoadsorption
- Weak evidence-base; benefit has been demonstrated for prolonged versus short duration treatment cycles
1 / III / 327
- Duration of subsequent immunosuppressive therapy: at least three years (by extrapolating from RCTs and retrospective cohort data); in class II-IV LN cases treated with MMF, tapering of MMF dose <2 g/day before 18 months after renal remission has been associated with increased risk (HR 6.3-6.8) for relapse