Update: Treatment of systemic lupus erythematosus


  • Donato Alarcón-Segovia

    Corresponding author
    1. Instituto Nacional de Ciencias Medicas y Nutricion Salvador Zubiran, Mexico City, Mexico
    • Instituto NAL Nutricion, Vasco de Quiroga #15 Mexico City, 14000 Mexico
    Search for more papers by this author

The current goal of treatment in systemic lupus erythematosus (SLE) is treatment-free remission. This is obtained by conventional therapy in 25% of patients, and many patients who achieve remission remain without disease activity for many years. Thus, long-term SLE survivors have a high prevalence of treatment-free remission (1). Current conventional therapy for SLE and its complications, together with advances leading to earlier diagnosis, have resulted in markedly improved patient survival rates. The situation, however, is still much less than ideal. Corticosteroids and immunosuppressors, the current mainstay of SLE treatment, cause increased morbidity and late-stage mortality, even when SLE may have become quiescent. In addition, the occurrence of lupus nephritis predominantly in young women may prevent the use of cyclophosphamide, because of the potential for infertility due to ovarian atrophy.

Knowledge of the pathogenesis of SLE has permitted new forms of treatment to be devised. One represents an interesting alternative to the use of cyclophosphamide in lupus nephritis (2). Another utilizes the synchronic effects of pulse cyclophosphamide and methylprednisolone (3). Others have reached pilot (4), I, or early II (5, 6), trial phases, and some may soon be openly reaching patients, particularly if they attain fast-track status for Food and Drug Administration approval. There are still some treatments in the experimental stage utilizing murine models of SLE that are potentially exciting, but discussing them would be beyond the scope of this update. Readers are referred to a recent review on the future of treatment for SLE (7).

Immunosuppressor options for lupus nephritis

Patients with diffuse proliferative lupus nephritis tend to do poorly when receiving glucocorticoids alone, and therefore require the addition of immunosuppressors. However, there are few immunosuppressor regimes available for them. If one studies the reports from the National Institutes of Health describing treatment with intravenous (IV) pulse cyclophosphamide (CPP), one finds that patients given prednisone, azathioprine, and small-dose oral CPP did practically as well as those patients receiving IV pulse CPP. Some patients with milder forms of SLE may achieve control of nephritis with only prednisone and azathioprine, but the majority of patients do require the addition of CPP to their regimens. However, the caveat of potential sterilization by CPP may prevent its use in many patients. New immunosuppressors without this morbidity but with similar effectiveness are therefore welcome.

Oral mycophenolate mofetil may be as effective an immunosuppressive agent for diffuse proliferative lupus nephritis as cyclophosphamide (New Eng J Med, 2000) (2)

Mycophenolic acid, an active metabolite of mycophenolate mofetil (MPM) is a purine nucleotide synthesis inhibitor that depletes lymphocytes and monocytes of guanosine triphosphate. It suppresses lymphocyte proliferation, decreases antibody formation, and interferes with the glycosylation of adhesion molecules. It seems to be a more effective immunosuppressor than azathioprine (AZA) appears less toxic than CPP, and also has no mutagenic effects.

For this study the authors randomly assigned, within 48 hours of biopsy, 42 SLE patients with diffuse proliferative glomerulonephritis to a group receiving MPM and prednisolone, or CPP (followed by AZA after 6 months) and prednisolone. The initial dose of prednisolone was the same for both groups (0.8 mg/kg/day). It was reduced by 5 mg/day every 2 weeks until a dose of 20 mg/day was reached, and by 2.5 mg/day every 2 weeks for 4 weeks and every 4 weeks thereafter until a maintenance dose of 10 mg/day was reached. The initial dose of MPM was 1 gm twice a day and was cut in half at 6 months. The dose of CPP was 2.5 mg/kg/day, and after 6 months it was changed to AZA (1.5 mg/kg/day). The study lasted 12 months, and after this time both groups of patients continued taking AZA 1 mg/kg/day. The primary endpoint of the study was remission, defined by the authors as proteinuria of <0.3 gm/24 hours, normal urinary sediment, normal serum albumin, and values of serum creatinine and creatinine clearance that were 15% or less above the baseline values.

At 8 weeks, both groups had improved similarly, although the decrease in serum creatinine concentration occurred faster in patients taking CPP than in those taking MPM. However, at 12 months creatinine clearance did not differ significantly from baseline values in either group. Achievement of complete (or partial) remission and the length of time on treatment it took for remission to occur were similar in both groups. Of the 39 patients who had a complete or partial remission, 5 had a relapse after 9 months on maintenance treatment (3 taking MPM, and 2 taking CPP/AZA).

Patients taking CPP/AZA (n = 7) had more infections than those receiving MPM (n = 4). Three patients taking CPP/AZA, but none receiving MPM, developed amenorrhea. Other adverse effects (except diarrhea) were also more frequent in the CPP/AZA group, although not significantly so. Because the comparison was made with oral rather than with IV CPP, assumptions on side effects with the more commonly used IV pulse regime cannot be made. However, both the results achieved and the paucity of side effects on MPM makes it an interesting alternative treatment for diffuse proliferative lupus nephritis.

Combination therapy with pulse cyclophosphamide plus pulse methylprednisolone improves long-term renal outcome without adding toxicity in patients with lupus nephritis (Ann Intern Med, 2001) (3)

The authors had previously shown that combination pulse therapy with IV CPP and methylprednisolone was better than methylprednisolone alone, and could result in a higher rate of renal response and fewer relapses than treatment with CPP alone. The two problems with this initial evaluation were the size of the cohort (n = 82) that, when divided into 3 treatment groups, was left with little statistical power, and the length of followup (median 5 years). The latter is particularly problematic when evaluating side effects because aseptic necrosis of bone and osteoporosis may appear much later. The followup has now been increased to a median of 11 years, with data on all 65 patients who had completed the 5-year protocol.

The first caveat continues to be valid because they still find no difference between the 3 treatment groups for risk of death or end-stage renal disease in an intent-to-treat analysis. When they use a composite endpoint of death, progression of renal disease, or need for additional immunosuppression, they find that patients receiving CPP or combination therapy were significantly less likely to experience treatment failure than those receiving methylprednisolone alone. Although the CPP and combination therapy groups did not differ statistically in this analysis, fewer patients in the combination group reached the composite endpoint (relative risk 0.59, 95% confidence interval 0.29–1.2). This would still be related to the problem of the limited size of their cohort. Fewer patients in the CPP group developed aseptic necrosis of bone during the treatment protocol, but in the extended followup time all 3 groups had a similar occurrence. This may have been due to 2 factors: 1) all patients had magnetic resonance imaging scans of coxofemoral joints at this time, and 2) all patients received high doses of oral prednisone at the outset, and were therefore at risk of developing aseptic necrosis in the long run. No differences were found in the frequency of osteoporosis, although the authors do not mention osteopenia. The other side effects in these patients are infections, and it seems that the main factor is CPP pulse therapy with little added influence from the methylprednisolone pulses.

Although the authors promote the notion that combination therapy of pulse CPP and pulse methylprednisolone should be widely used, I would still reserve its use for patients whose renal biopsies show high activity indices, including evidence of potential accelerated progression (e.g., extra-capillary proliferation).

Interventions at specific points of the immunoregulatory circuits

Immunosuppressors such as CPP, AZA, or MPM act almost indiscriminately on the immune system and therefore cause a certain degree of immune deficiency that is paid for by liability to infections. On the other hand, it makes them effective for use in various autoimmune diseases and in organ transplantation. However, if we are going to achieve a cure or at least induce remission in most patients without the high cost of morbity or mortality, more rational forms of treatment are imperative. Many years ago, we postulated that knowledge of the immunoregulatory circuits in each of the connective tissue diseases was needed for this purpose (8). We found that each disease reached autoimmunity by different pathways and thus their treatment would be different. This has resulted in opposing treatment possibilities for SLE and rheumatoid arthritis (RA). Anti-tumor necrosis factor therapies are now available for RA patients both with monoclonal antibodies (mAb) or recombinant receptor ligands. This may not be appropriate for SLE, but trials with mAb for SLE intervention at different immunoregulatory levels have now taken place and can teach us several lessons.

Intervention against the CD40-ligand is feasible in SLE patients (J Rheumatol, 2001) (5)

The activation of T cells, permitting them to stimulate other effector cells, is at least a two-step process. The first step is achieved by interaction between the T cell receptor and the antigenic peptides prepared and presented to it by cells of monocytic lineage known as antigen presenting cells (APC). The second step consists in T cell costimulation provided by signals involving receptor–ligand pairs on T cells or on APC. One of these entails interaction between the CD40 molecule on APC and its ligand (CD40L) on helper (CD4+) T cells. Also, the interaction between CD40L on T cells and CD40 on B cells provides extreme help to these, causing them to activate, proliferate, respond to cytokines, produce IgG antibodies, become resistant to apoptosis, and induce costimulation of other B cells. Patients with SLE may have increased numbers and expression of CD40L+ T cells, as well as elevated levels of soluble serum CD40L. In addition, the expression of CD40L on T cells from SLE patients may be prolonged, and this may contribute to the dysregulation. Intervention at the level of CD40 to CD40L interaction has been found beneficial in murine models of lupus.

In the study, symptomatic SLE patients received a single dose of a humanized monoclonal antibody to CD40L (IDEC-131) that has been shown to bind both avidly and specifically to CD40L on T cells and thus prevent the costimulatory signal. The single dose of IDEC-131 was escalated in cohorts of 3 to 5 patients each, from 0.05, 0.25, 1.0, 5.0, up to 15.0 mg/kg, and administered in a 2-hour IV infusion. Patients could be receiving a stable dose of prednisone, hydroxychloroquine, methotrexate, or nonsteroidal antiinflammatory drugs, but not other immunosuppressors or other biologic agents.

The surface antigen expression of CD3, CD4, CD8, and CD40L T cells and of CD19 B cells was determined by flow cytometry. The pharmacokinetics of the mAb were determined by enzyme-linked immunosorbent assay using the extracellular domain of CD40L linked to CD8 as antigen.

A total of 23 SLE patients with mild to moderate clinical manifestations were enrolled and followed for 3 months. Main adverse events were mild nausea, dizziness, and headache. There was no evidence of hepatic, renal, or hematologic toxicity nor of infections. There was also no evidence of treatment-related T or B cell depletion nor were there detectable antibodies to IDEC-131 in any patient. The pharmacokinetics of the higher dose groups were similar (half-life mean range: 298.7–319.7 hour) but the 0.25 mg/kg dose group had shorter half-life (123.0). Although the single dose did not result in any noticeably favorable clinical effect, the study did pave the ground for a phase II study of the effect of this form of treatment.

Preliminary results were presented at the 64th Annual American College of Rheumatology (ACR) meeting, Philadelphia, PA, October 2000 (9). The trial included 85 patients divided in 4 groups and treated with either saline or one of the 3 higher doses of IDEC-131. Each received 6 infusions over 16 weeks and had as primary efficacy endpoint the Safety of Estrogens in Lupus Erythematosus: National Assessment and Systemic Lupus Erythematosus Disease Activity Index score at 20 weeks. There were no differences in either efficacy nor in adverse events.

Another study was presented at the same meeting using a different anti-CD40L mAb (hu5c8) on 5 SLE patients with nephritis (10). A 2- to 8-fold decrease in IgG and IgM secreting B cells was seen, along with a 10- to 200-fold decrease in IgG and IgM anti-DNA secreting B cells that in 4 of 5 patients became undetectable. These reverted 168 days after the last treatment but remained below pretreatment levels in 2 patients. It would therefore seem that intervention at the CD40L could have profound and persistent effects on peripheral blood B cells and in particular on those that produce anti-DNA.

It would appear that intervention is feasible at the costimulatory level of immune regulation in SLE patients, but that this approach should be refined both in regard to the mAb used, as well as potential efficacy, perhaps with different patient selection.

Clinical and biologic effects of anti-interleukin-10 monoclonal antibody administration in systemic lupus erythematosus (Arthritis Rheum, 2000) (4)

Monocytes and B cells from SLE patients produce abnormally large amounts of interleukin-10 (IL-10), and serum levels of IL-10 correlate with disease activity. SCID mice injected with mononuclear cells (MNC) from SLE patients develop anti-DNA antibodies that are virtually abolished by anti-IL-10 mAb, but not by anti-IL-6 injection. Increased production of IL-10 explains most of the immunoregulatory defects found in SLE including the decreased production of, and response to, IL-2 and IL-1 with increased production of Th2 cytokines and impaired cell-mediated immune responses. A pilot study was therefore conducted in 6 steroid-dependent SLE patients who were given a 21-day course of 20 mg/day of an anti-IL-10 murine mAb (B-N10) and followed for 6 months. The treatment caused no untoward effects except for chills during mAb infusion on day 16 in one patient. It was attributed to immunization against the mAb and further treatment was discontinued.

A steady-state plasma level of 20 μg/ml of B-N10 was observed on days 9–11. On day 30, B-N10 was detectable in only 2 patients but by day 60 had disappeared in both. By monitoring the plasma concentration of IL-10 it could be determined by the ratio of B-N10 to IL-10 that the dose of mAb was sufficient to neutralize the IL-10 being produced. All patients had developed anti-mouse antibodies before day 60. The Mexican modification of the Systemic Lupus Erythematosus Disease Activity Index scores had decreased significantly in all patients by day 21 and decreased further by day 60. They remained low for 6 months. This allowed significant decreases in prednisone requirements by the first month that declined further up to the end of the study. The anti-DNA antibody levels decreased in only one patient but there were no changes in C3 or C4 levels.

The study includes information on the acute effect of anti-IL-10 administration on several immunologic abnormalities. Comparison was made between day 1 and day 10 samples as there had been no changes in prednisone dose in any of the patients by this time. There were decreases in T lymphocyte activation markers IL-2R, in a marker of general immune cell activation (sTNFR p75), and in several markers of endothelial cell activation (sICAM-1, sVCAM-1, and vWF). The spontaneous releases of sIL-2R and IL-10 also decreased significantly. In contrast, there were increases in the serum concentration of IL-1Ra, an antiinflammatory cytokine, and of IL-2 in the supernatants of PHA-stimulated MNC. All these findings indicate a prompt decrease of immune activity and partial restoration of T lymphocyte function upon administration of the anti-IL-10 mAb. Because of the appearance of human anti-mouse antibodies, repeated administration of this non-humanized mAb was not feasible. The study supports the central role of IL-10 production in the immune dysregulation of SLE and indicates a pathway of treatment that may be followed either with a humanized mAb or with drugs that abrogate IL-10 production (11).

Recovery of tolerance

Autoimmunity is a natural and physiologic process. However, autoimmune disease implies loss of immune tolerance with auto reactive T cells and somatically mutated B cells that produce pathogenic autoantibodies. There are several ways of inducing tolerance to autoantigens. By administering an autoantigen via a tolerizing route, be it mucosal or dermal, or by giving it in tolerizing form (e.g., soluble IV or intraperitoneal, alone or coupled to MHC peptides, as an altered peptide ligand, or as an aggregated protein chimera) (7). It can also be achieved by administering antigen-specific T cells in an immunogenic form.

Oral tolerance has been attempted for RA by ingestion of chicken collagen. It has also been attempted in NZB × NZW F1 mice orally given a rat-kidney extract that would include the putative kidney antigen of nephritogenic SLE autoantibodies (12). This form of treatment resulted in decreased anti–double-stranded (ds) DNA antibodies, reduced kidney damage, and prolonged survival as compared to controls. A recent development for seeking specific B cell tolerance to DNA consists of a triethylene glycol plate coupled to 4 dsDNA epitopes of 20 base pairs each. The compound, named LJP-394, with a molecular weight of 54 kD would bind to the anti-dsDNA surface immunoglobulin of anti-dsDNA producing B cells without T cell help, thus resulting in their apoptosis.

Treatment of systemic lupus erythematosus with LJP-394 (J Rheumatol, 2001) (6)

Fifty-eight SLE patients were randomly assigned to receive 1, 10, or 50 mg of LJP-394 or placebo by IV injection. After a 2-month pretreatment period, the drug or placebo were started also randomly administered at weekly, biweekly, or monthly intervals for a total of 16 weeks. These were followed by a 2-month posttreatment period. The study revealed dose-related decreases in anti-DNA antibodies caused by LJP-394. Thus, patients who received 50 mg of LJP-394 had reduction of 38.1% and 37.1% at weeks 16 and 24, respectively. Little or no decrease was noted on the 1 mg dose or in any doses given monthly or biweekly. Weekly administration of 10 mg of LJP-394 resulted in a 29.3% decrease of anti-DNA titers at week 24. Adverse effects were comparable in the placebo and LJP-394 treatment group.

This study has been followed by 2 reports. One showed that specific affinity of the anti-DNA for the DNA moieties of LJP-394 at pretreatment was higher than after treatment with 10 or 50 mg of LJP-394 weekly. This suggests a fall of a subgroup of anti-DNA antibodies with high affinity for the 20 base-pair oligonucleotide present in this drug (13). Another study presented at the 64th Annual ACR meeting, Philadelphia, PA, October 2000 studied the potential prevention of renal flares with the use of this compound compared with placebo (14). In this trial, 211 SLE patients with anti-DNA antibodies and a history of renal flare received weekly infusions of 100 mg of LJP-394 or placebo for 16 weeks followed by intermittent dosing with LJP-394 or placebo for 60 weeks. The primary endpoint was renal flare indicated by a >1 gm increase in proteinuria, a two-fold increase in protein in two 24 hour urine collection, a reproducible 20% increase in serum creatinine, or a 2 grade in hematuria with dysmorphic cells. The trial was stopped when an interim analysis showed 19 renal flares in the drug group and 23 in the placebo group. However, when the data were analyzed by the pretreatment affinity of anti-DNA antibodies to the LJP-394 moieties, the time to renal flare was significantly longer in drug-treated patients with high affinity antibodies. It would therefore appear that SLE patients with high affinity antibodies to the oligonucleotides present in LJP-394 could benefit from their administration. Fortunately, these were present in 88% of SLE patients selected for the study.

Although the prognosis of SLE has improved remarkably in the last 2 decades, there are still refinements necessary to currently available forms of treatment. A host of new forms of more effective treatments are looming on the horizon.