Chapter 32. Lupus Nephritis

• Antinuclear and anti-DNA antibodies.
• Depressed complement components, C3 and C4.
• Deposition of nephritogenic immune complexes.
• Spectrum of hematuria, proteinuria, hypertension, and renal insufficiency.
• Major pathologic classes of lupus nephritis (LN): Class I, minimal mesangial LN; Class II, mesangial proliferative LN; Class III, focal LN; Class IV, diffuse LN (segmental/global); Class V, membranous LN; and Class VI, advanced sclerosis LN.
• Variable acute and chronic renal vasculopathy and tubulointerstitial damage.

Evidence of pathogenic autoimmunity within the kidney is manifest in the majority of patients with systemic lupus erythematosus (SLE). The clinical and pathologic effects of the disordered autoimmune processes range from mild and indolent to severe and progressive glomerulonephritis.

The immunologic basis for autoimmunity in SLE is complex and incompletely understood, but there is evidence of dysregulation of both humeral and cellular immune pathways. Autoantibodies to a range of constitutive nuclear antigens are cardinal elements in the definition of SLE. Of these, anti-DNA antibodies appear to be the most nephritogenic, although other autoantibodies are clearly involved in many cases.

Immune complexes of several types appear to be involved. The classic paradigm involves deposition of preformed circulating immune complexes, initially in the mesangial interstices with eventual spillover into the subendothelial space manifested by progressive stages of mesangial and endocapillary proliferative disease. An alternative model posits that circulating nuclear remnants of excessive apoptotic cellular breakdown are bound to glomerular capillary sites with subsequent binding of autoantibodies to the “planted” antigens and local formation of immune complex aggregates.

Observations in the Heymann nephritis animal model and in rare congenital forms of human membranous nephropathy indicate that autoantibodies are capable of reacting with constitutive antigens of the glomerular podocytes. This in situ production and accumulation of immune complexes along the external (subepithelial) aspect of the glomerular capillary has not been definitively proven to be operant in human membranous LN.

Autoimmunity based on direct cell-mediated injury to various components of the nephron have been proposed, based primarily on the high frequency of mononuclear cell infiltration in various forms of LN. Despite the common presence of infiltrating lymphoid cells, there is no direct proof of a pathogenic mechanism (analogous to the nephron damage of cell-mediated renal allograft rejection) in human LN.

The natural history of SLE is known most precisely from studies of murine models where death from progressive LN regularly occurs unless interdicted by effective immunosuppressive regimens. Evidence of a comparably ominous natural history of human LN is occasionally evidenced by severe damage and renal failure in misfortunate patients lacking access to modern medical therapies. It is gratifying that numerous therapeutic advances have dramatically improved the prognosis of LN over the past few decades.

The clinical manifestations of SLE are extremely protean. In the context of clinical practice, diagnosis of SLE should be suspected when there is evidence of multisystemic disease along with autoantibodies to nuclear antigens. Ideally, the diagnosis of SLE should be based on the presence of four or more criteria defined by the American College of Rheumatology, but it should be emphasized that these criteria are intended to standardize eligibility criteria for inclusion of subjects in SLE research studies, rather than to establish diagnostic criteria for SLE in clinical practice.

Symptoms and Signs

Glomerulonephritis is uncommonly the sentinel manifestation of SLE. An exception to this principle is membranous LN wherein up to 25% of patients present with no extrarenal manifestations and diagnosis of SLE may emerge only during extended follow-up.

The key challenge for diagnosticians is to detect clinically significant LN prior to overt symptoms. Patients with a likely or proven diagnosis of SLE should be carefully queried for what they may ordinarily consider to be trivial changes in urine color, nocturia, and foam-producing urination, each of which may mark the onset of occult LN.

Laboratory Findings

Evidence of hematuria, proteinuria, or pathologic urine sediment on screening urinalysis usually confirms the presence of LN. However, the astute clinician should be mindful that false-negative urinalyses, particularly urine sediment abnormalities, are relatively common in high-throughput clinical laboratories. It is judicious to “flag” urinalyses from patients with SLE for special scrutiny by laboratory personnel.

Proteinuria is traditionally quantified by 24-hour urine collection. However, spot urine protein/creatinine ratio (the value of which approximates the number of grams per day of proteinuria) is becoming widely accepted as the more convenient alternative for patients, particularly for monitoring changes in proteinuria over time.

Serologic testing includes antinuclear, anti-DNA, and antiphospholipid antibodies, which are useful for the diagnosis of SLE and its complications. The clotting diathesis conferred by the antiphospholipid syndrome may compound the thromboembolic risk of persistent nephrotic syndrome. Elevated anti-DNA titers and depressed levels of C3 and C4 complement levels correlate with active proliferative LN; changes in levels of these serologies tend to be more useful than their absolute levels for monitoring activity of nephritis during follow-up.

Special Tests

Renal biopsy is primarily valuable for staging the type and severity of LN. The classification scheme has recently been revised and updated (Table 32–1). Discordances among clinicopathologic findings are common over time and repeat renal biopsy may be needed to stage (or potentially reclassify) the renal disease.

Knowledge of the findings of renal biopsy pathology strongly influences therapeutic decisions in LN. Institution of intensive therapy, particularly cytotoxic drugs, has been shown to occur earlier when clinicians have the benefit of renal pathology data.

Several systemic diseases, particularly the systemic vasculitides, have features that commonly overlap those of SLE and LN. Antineutrophil cytoplasmic antibodies (ANCA) usually allow discrimination between LN and renal vasculitis, but it is necessary to be mindful that a small subset of patients with a definitive diagnosis of SLE have ANCA autoantibodies.

Patients with rheumatoid arthritis may have hematuria in association with mild mesangial nephritis or proteinuria in association with membranous nephropathy. Mixed connective tissue disease has been infrequently associated with glomerulonephritis, which, however, may be indistinguishable from the various classes of LN.

Thrombotic microangiopathy is known to occur as a superimposed process in some patients with SLE and LN. The potential role of anticardiolipin or antiphospholipid antibodies as a cause of lupus kidney disease is unresolved, but it seems to account for some cases of extraglomerular vasculopathy in patients with SLE.

The complications of LN result from the pathophysiology intrinsic to the glomerular disease, including hypertension, nephritic and nephrotic syndromes, and renal failure, as well as from side effects of treatment. While clinicians treating patients with LN have traditionally focused on therapeutic interventions to reduce the risks of renal failure, there is emerging appreciation that treatment may be required to interdict the cardiovascular and thromboembolic complications engendered by protracted nephrotic syndrome. Indeed, evidence of the benefit of achieving even partial remission of proteinuria (to the subnephrotic level) has a salutary effect on patient and renal survival. Beyond standard immunosuppressive therapies, the full armamentarium of renal protection strategies, particularly angiotensin antagonists and lipid-lowering statin drugs, is warranted in management of LN.

Optimal care of patients with LN usually requires integrated expertise of nephrologists and rheumatologists. Most SLE patients will require some dosage of corticosteroids, antimalarials, and nonsteroidal anti-inflammatory drugs for control of their commonly debilitating extrarenal disease, which are best evaluated and managed by rheumatologists. Conversely, delineation of the more arcane aspects of renal disease and integration of the results of renal biopsy are best evaluated and managed by nephrologists. Conjointly staffed clinics offer the best environment for effective communication and comprehensive care of patients with SLE and LN.

Immunosuppressive drug options for management of patients with LN are summarized in Table 32–2. Evidence-based clinical recommendations derived from completed controlled clinical trials are limited and consensus in developing clinical practice guidelines has not been achieved. Results of ongoing multicentered clinical trials are expected to help prioritize the current therapeutic options for treatment of the various forms of LN within the next few years.

Corticosteroids

Patients with new onset Class III, IV, or V LN warrant a limited therapeutic trial of high-dose corticosteroids. Patients failing to achieve full remission of nephritis within 6–8 weeks should be treated with adjunctive cytotoxic drug therapy (cyclophosphamide or mycophenolate mofetil) as described below. Based on current evidence, those patients with substantial fibrinoid necrosis or cellular crescents on renal biopsy should be directly initiated on therapy with combined pulse methylprednisolone and pulse cyclophosphamide for a period of 6 or more months.

Cyclophosphamide

Results of early treatment of murine LN and meta-analyses of human trials indicate that cyclophosphamide is among the most effective immunosuppressive drugs for LN. Side effects of daily cyclophosphamide are formidable, particularly beyond 3 months, and for this reason, prescription of daily therapy has become limited. The therapeutic index of cyclophosphamide is improved by administration of intermittent pulse therapy, which has become widely accepted as the standard approach to administration of cyclophosphamide therapy.

The emergence of pulse cyclophosphamide as the therapeutic standard for proliferative LN arose from observations in several long-term clinical trials reported from the National Institutes of Health. The composite of evidence indicated that pulse cyclophosphamide achieved the most sustained rates of renal remission, the lowest rates of cumulative damage on renal biopsy, and ultimately the lowest rate of progression to end-stage renal failure. The substantive risk of gonadal toxicity incurred with extended courses of pulse cyclophosphamide has been the major impetus to continue the search for alternative therapies in LN.

Mycophenolate Mofetil

Mycophenolic acid was initially tested and shortly abandoned for treatment of rheumatic diseases in the 1960s. It was reformulated and initially compared to azathioprine in renal allograft recipients. Mycophenolate was licensed on the basis of its advantage over azathioprine in achieving a reduced frequency of acute rejection episodes. Doubts have been raised about the cost effectiveness of mycophenolate because of its high cost and its lack of proven benefit in extending allograft survival.

Several uncontrolled trials of mycophenolate in LN suggested that this therapy may be useful in patients failing to achieve satisfactory responses to cyclophosphamide. Since 2000, two controlled trials comparing induction therapy with mycophenolate mofetil versus cyclophosphamide (Chan study: Daily cyclophosphamide and Ginzler study: Monthly pulse therapy) indicated comparable rates of renal remission and short-term renal survival but fewer side effects in patients treated with 1–3 g/day of mycophenolate mofetil. A controlled trial led by Contreras comparing maintenance therapies with cyclophosphamide, azathioprine, and mycophenolate mofetil indicated that the best patient and renal outcomes were achieved with mycophenolate.

Because historical studies have shown that several treatments, including azathioprine, cyclophosphamide, a combination of azathioprine and cyclophosphamide, and even prednisone alone, achieve comparable short- and intermediate-term renal survival rates, many have argued that the definitive choice among the newer immunosuppressive drugs must await ascertainment of long-term renal outcomes after 5 or more years of observation.

Azathioprine

Azathioprine is a relatively weak immunosuppressive drug in studies of both murine and human LN. It has mostly been relegated to use as a steroid-sparing agent and as a less costly and well-tolerated maintenance after cyclophosphamide or mycophenolate induction therapies.

Experimental Therapies

Several studies involving novel therapeutic agents for LN are underway. The agents being evaluated in these studies are summarized in Table 32–3. The Web site www.clinicaltrials.gov is a potentially useful resource in searching for studies that may be recruiting patients, along with information about eligibility and exclusion criteria.

Membranous Lupus Nephritis with Stable Renal Function

Patients with membranous LN with normal renal function and subnephrotic proteinuria may not warrant aggressive immunosuppressive therapy and may be adequately managed with angiotensin antagonists and statins as needed for hyperlipidemia. Those with high-grade nephrotic-range proteinuria, particularly if protracted and unimproved by angiotensin antagonists, should be treated with similar but less intense regimens of corticosteroids plus cyclophosphamide or mycophenolate. If renal function is well preserved, cyclosporine is also an effective alternative in patients with membranous LN. The optimal duration of treatment is undefined, but relapse of proteinuria is particularly likely after cessation of cyclosporine. This has prompted a very protracted slow tapering of cyclosporine before discontinuation (unless there is clinical suspicion or pathologic evidence of significant cyclosporine nephrotoxicity).

The prognosis of Class III and IV proliferative LN has improved from a 5-year renal survival of <20% during the period 1960–1980 to >80% during the period 1980–2000. This improvement in prognosis has been ascribed mostly to increasing use of cyclophosphamide. While preliminary data based on achievement of renal remission suggest that mycophenolate may have comparable benefits, it remains to be established if mycophenolate will achieve comparable long-term renal survival.

In patients progressing to end-stage renal disease due to LN, there is considerable controversy about the prevalence of active SLE during maintenance dialysis and the risk of recurrent LN in renal allografts. In general, patients should be clinically and serologically inactive for approximately 1 year before renal transplantation. Recent reports suggest that recurrence of low-grade LN is common, but fortunately clinically significant nephritis and allograft loss are rare.