SEARCH

SEARCH BY CITATION

Introduction

  1. Top of page
  2. Introduction
  3. Difficulty studying lupus
  4. Trial designs in lupus
  5. Definitions of surrogate marker and biomarker
  6. Potential roles and uses of biomarkers
  7. Examples of potential surrogates and biomarkers
  8. Difficulties with the use of biomarkers/surrogates in lupus trials
  9. Future approaches
  10. Conclusion
  11. REFERENCES

Clinicians and representatives of academia, industry, the National Institutes of Health (NIH; both intramural and extramural programs), and the Food and Drug Administration (FDA) met March 31–April 1, 2003 in Bethesda, MD to discuss the possible uses of biomarkers and surrogate markers in the design of trials to study new therapies for systemic lupus erythematosus (SLE). This report summarizes the intricacies in studying a disease such as lupus and recounts discussions regarding the potential uses of biomarkers, their possible role in lupus trials, and some of the types of biomarkers that might be used in these trials. Finally, there will be a discussion of some of the pitfalls and problems associated with the use of biomarkers and surrogate markers, along with some possible solutions.

Difficulty studying lupus

  1. Top of page
  2. Introduction
  3. Difficulty studying lupus
  4. Trial designs in lupus
  5. Definitions of surrogate marker and biomarker
  6. Potential roles and uses of biomarkers
  7. Examples of potential surrogates and biomarkers
  8. Difficulties with the use of biomarkers/surrogates in lupus trials
  9. Future approaches
  10. Conclusion
  11. REFERENCES

At present, only 3 drugs have been approved for use in the treatment of SLE. These include hydroxychloroquine, glucocorticoids, and low-dose aspirin. In the last 3 decades, no new drugs have been approved for the treatment of lupus. This underscores the fact that lupus, due to a number of factors, is a challenging disease to study. These factors include the heterogeneity of lupus and a clinical course characterized by flares and remissions that are difficult to predict. Hence, the progression of the disease is quite variable. Often, the time to occurrence of measurable, important, clinically signifi cant change is prolonged, and recurrences or flares are difficult to anticipate. Because of this variability, long-term trials to demonstrate clinical benefit may be necessary. The unpredictable nature of the disease is further supported by the fact that at least 5 years of followup in the NIH lupus nephritis trials were required to ascertain a sign of efficacy of cyclophosphamide in preventing renal deterioration in patients with active lupus nephritis (1). Contrast this with diabetic nephropathy, for example, in which deterioration of renal function can be predicted with confidence.

Another challenge in the study of lupus is the lack of clearly defined and well-validated outcome measures that are sensitive to change in disease status while meaningful in terms of improvement of survival and quality of life. Furthermore, studies need to take into account the potential for “fixed” damage induced not just by the disease, but also by treatment. Additionally, the optimal length of time to determine the appropriate outcomes for study is not clear. To complicate matters further, morbidity and mortality due, for example, to renal disease may have been spontaneously improving over the last 3 decades. Patterns of morbidity are changing, especially as related to cardiovascular disease. This overall change in mortality and morbidity may be multifactorial and related to disease activity as well as to improved treatment options.

All of the above issues point to the difficulties in performing definitive trials in lupus. These difficulties emphasize the point that well-designed and well-implemented trials in lupus will likely come with high costs. Furthermore, the performance of clinical trials in lupus would benefit by the ability to recruit adequate numbers of patients and to allow for use of standardized efficacy and safety measures.

Trial designs in lupus

  1. Top of page
  2. Introduction
  3. Difficulty studying lupus
  4. Trial designs in lupus
  5. Definitions of surrogate marker and biomarker
  6. Potential roles and uses of biomarkers
  7. Examples of potential surrogates and biomarkers
  8. Difficulties with the use of biomarkers/surrogates in lupus trials
  9. Future approaches
  10. Conclusion
  11. REFERENCES

At the conference, there was a discussion of trial design, and several possible approaches were presented. Two broad categories of studies were proposed. Investigators in the first type of trial would study the rate of flare or time to flare in stable patients with inactive lupus. In the second design category, trials of lupus patients with active disease would utilize end points of either time to remission or total glucocorticoid dose used (thus, “steroid sparing” trials). Additional discussed aspects of study design included the use of open-label studies, comparison against placebo, and comparison against active comparator (either superiority or noninferiority in design). Other approaches discussed included add-on trials, crossover trials, trials with historical controls, or withdrawal-of-active-treatment trials, each of which has advantages and limitations.

The issue of placebo-controlled trials was considered in detail. The discussion centered around add-on trials using “standard of care” to which a new drug or placebo is added. Regarding trials of treatment of active lupus nephritis, the question was raised as to whether the “standard of care” could be just glucocorticoids alone or steroids plus cyclophosphamide. The utility, practicality, and ethics of add-on trials using standard of care must be balanced against the fundamental need to understand the characteristics of the new drug by itself in terms of toxicity and efficacy, particularly for trials designed for regulatory actions. The add-on design makes analysis of toxicity related to the new therapy difficult, because in this model drugs are used in combination. It was proposed that glucocorticoids alone may be appropriate, at least in the short term, for the initial treatment of lupus nephritis in controlled trials of a new therapy. It appears, however, that institutional review boards have been resistant to this approach because cytotoxic therapy is considered part of the standard of care for lupus nephritis.

A somewhat analogous situation exists in the study of rheumatoid arthritis (RA), in which new therapies are studied as add-on treatment with disease-modifying antirheumatic drugs (DMARDs) such as methotrexate, and in patients who are considered partial DMARD responders. In this case, not only does the add-on design pose a problem for identifying specific issues related to the new drug (unless the new drug is studied later as monotherapy), but it also poses a problem for defining the characteristics of a “partial responder.” Therefore, new therapies studied only in add-on trials become combination therapies by default, because they are never studied alone. Furthermore, drug–drug interactions may produce additional untoward adverse events which cannot be attributed to any one drug if only add-on trials are used to establish efficacy and safety of a new drug.

An alternate design considered at the conference was that of noninferiority trials with cyclophosphamide as the comparator. Noninferiority trials are defined as trials in which the effect of a new therapy is compared with that of another active therapy without the benefit of a placebo control group. The new therapy must maintain a predefined percentage of the active comparator's effect in order to be considered noninferior to the comparator therapy. Such trials may be difficult to analyze from a regulatory point of view, since an effect size has not been established for cyclophosphamide alone. It may be possible, however, to determine an effect size for cyclophosphamide using data from meta-analyses (2) in which cyclophosphamide plus glucocorticoids is compared with placebo plus glucocorticoids. There is little prospective evidence of cyclophosphamide-only efficacy compared with placebo, and there are few data that clearly delineate appropriate doses; also, as noted above, effects on renal function may not be apparent for years. Furthermore, the effect sizes of various interventions using cyclophosphamide as the comparator are likely to vary depending on the organ involved as well as on the severity and extent of existing organ damage in patients enrolled in interventional trials.

Outcome measures.

Outcome measures discussed at the conference included organ-specific outcomes, or, alternatively, time to or frequency of flare, prednisone dose, responder index, or change in disease activity measure (such as the SLE Disease Activity Index [SLEDAI], Systemic Lupus Activity Measure [SLAM], British Isles Lupus Assessment Group [BILAG], European Consensus Lupus Activity Measure, or Lupus Activity Criteria Count [3–7]). No specific recommendations were made regarding whether trials should focus on organ-specific end points or disease activity end points.

There are advantages and disadvantages to each approach. For example, use of an organ-specific approach allows for a trial with a more homogeneous population, although recruitment of adequate numbers of patients may be more difficult. On the other hand, use of a disease activity index allows for recruitment of a greater number of patients, but trials will invariably have a more heterogeneous population and may be affected by imbalances in disease manifestations between treatment groups. Therefore, disease activity indices do not necessarily reflect outcome or change in disease activity in any specific organ. Additional difficulties with using disease activity indices include variability in weighting of organ manifestations and variability in scoring items. In fact, only the BILAG was designed to reflect an appropriate need for change in treatment. There is also a poor correlation between patient and physician assessment of disease activity. However, it was the opinion of the OMERACT (Outcome Measures in Rheumatology Clinical Trials) group (8) that any trial in lupus should include the following measures: disease activity measure, damage index (such as the Systemic Lupus International Collaborating Clinics/American College of Rheumatology SLE Damage Index [9]), health-related quality of life (such as the Short Form 36 health survey [10]), adverse events, and economic costs.

Definitions of surrogate marker and biomarker

  1. Top of page
  2. Introduction
  3. Difficulty studying lupus
  4. Trial designs in lupus
  5. Definitions of surrogate marker and biomarker
  6. Potential roles and uses of biomarkers
  7. Examples of potential surrogates and biomarkers
  8. Difficulties with the use of biomarkers/surrogates in lupus trials
  9. Future approaches
  10. Conclusion
  11. REFERENCES

A discussion of potential definitions of surrogate markers and biomarkers followed. In the regulatory realm, a clinical end point is defined as a characteristic or variable that measures or shows how a patient feels, functions, or survives. A surrogate end point is defined as a marker intended to substitute for a clinically meaningful end point that measures directly how a patient feels, functions, or survives. Changes induced by a therapy on a surrogate end point are expected to reflect changes in a clinically meaningful end point. One potential area in which surrogate markers may be useful is when clearly validated and relevant outcomes are not measurable in the time frame of the proposed clinical trials (outcomes may take years to develop). A surrogate marker could be measured in serum or urine, or might be the results of an imaging study.

According to 21 CFR (Code of Federal Regulations; FDA regulations pertaining to new drug approvals) 314.500 subpart H, a new drug may be approved on the basis of a less than fully established surrogate that is “reasonably likely, based on epidemiologic, therapeutic, pathophysiologic, or other evidence to predict clinical benefit.” To be approved on this basis, a drug must 1) treat serious or life-threatening disease, 2) provide meaningful therapeutic benefit over existing treatments, and 3) be studied after approval to verify and describe its clinical benefit. If approval is to be pursued with the use of a surrogate, the surrogate should have been validated prior to the trials for registration. The sponsor of the trial would then need to commit to postmarketing studies to demonstrate clinical benefit.

On the other hand, biomarkers generally refer to any physiologic, pathologic, or structural property or activity that can be influenced by a drug. These biomarkers can be measured in tissue, cells, or fluids. While surrogate end points are candidate bases for drug approval, biomarkers do not necessarily have the same regulatory implication for drug approval, since they do not necessarily reflect clinical outcomes. Therefore, a biomarker is not necessarily a surrogate marker. Nonetheless, biomarkers may be useful for predicting patient response.

Potential roles and uses of biomarkers

  1. Top of page
  2. Introduction
  3. Difficulty studying lupus
  4. Trial designs in lupus
  5. Definitions of surrogate marker and biomarker
  6. Potential roles and uses of biomarkers
  7. Examples of potential surrogates and biomarkers
  8. Difficulties with the use of biomarkers/surrogates in lupus trials
  9. Future approaches
  10. Conclusion
  11. REFERENCES

A number of roles for biomarkers in clinical trials were proposed. These roles included the following: 1) improvement in efficiency and design of trials; 2) improved understanding of a drug's effects and of how to use it, from both efficacy and safety perspectives; 3) demonstrating efficacy in a new setting; 4) support for results of clinical trials; and 5) understanding SLE pathogenesis. Other potential roles of biomarkers include use in diagnosis, determining prognosis, and identifying remissions/relapses of disease. Each of these roles was discussed further during the meeting.

Improvement in efficiency and design of trials.

The use of biomarkers may allow an investigator to choose a dose range of a drug from both efficacy and safety viewpoints and to identify the appropriate titration regimen if necessary. Use of biomarkers may provide insight into the development of tolerance to a drug's effects and identify adverse effects due to drug withdrawal that need further study. The strength of biomarkers may be in their potential role in defining populations more or less likely to respond to therapy (leading to an increase in homogeneous populations; similar prognosis). One possible approach to defining responder populations is by using biomarkers to identify the presence of appropriate receptors or targets that need to be present for a drug to produce an effect. However, it is not known whether patients without these receptors or targets will (or will not) respond to drug treatment. Therefore, a concern is that this approach limits the populations studied and, hence, the generalizability of the study results. As a means of identifying a population that is likely to respond to a new drug, however, this approach has great appeal. Furthermore, another advantage to using biomarkers may be to reduce the numbers of patients needed to adequately identify a clinically significant effect. Biomarkers may reduce false-positive outcomes. Taken together, the use of biomarkers in early trials may increase the speed and confidence with which the investigator proceeds to perform pivotal trials.

Improved understanding of drug effect.

Biomarkers may identify reasons for subgroup differences, dose and dose interval, effects over time, and pharmacodynamic interactions.

Efficacy in new setting.

Biomarkers may be used to demonstrate reasons for efficacy or lack thereof in pediatric populations, for example, or in different ethnic groups. Biomarkers may help establish appropriate dosing and dosing intervals in subgroups where efficacy has already been established. The use of biomarkers may allow for bridging from adult efficacy to pediatric efficacy without additional full clinical studies.

Support for results of clinical studies.

The results of one clinical study where a clinical outcome is measured may be supported by another study in which biomarkers are used as an outcome. For example, it might be possible to demonstrate efficacy of a new therapy based on two studies. The results of one study might demonstrate a clinically important renal disease–related outcome (for example, the resolution of hematuria and proteinuria) that correlated with a change in anti–double-stranded DNA (anti-dsDNA) antibody levels, while a second study might demonstrate a significant change in anti-dsDNA antibody levels only.

Understanding SLE pathogenesis.

Biomarkers may help identify immunologic mechanisms leading to tissue damage. For example, changes in cytokine production or in expression levels of cytokine receptors on appropriate cell types might provide insights about the mechanisms involved in the pathogenesis of lupus.

Identifying remissions/relapses of disease.

Biomarkers may allow the early determination of which patients will benefit from changes in therapy (either increases or decreases in dosage) to prevent or treat various manifestations of the disease process. For example, if complement deposited on erythrocytes can predict future flares of disease, measurement may allow changes in therapy before a clinical flare occurs.

Examples of potential surrogates and biomarkers

  1. Top of page
  2. Introduction
  3. Difficulty studying lupus
  4. Trial designs in lupus
  5. Definitions of surrogate marker and biomarker
  6. Potential roles and uses of biomarkers
  7. Examples of potential surrogates and biomarkers
  8. Difficulties with the use of biomarkers/surrogates in lupus trials
  9. Future approaches
  10. Conclusion
  11. REFERENCES

None of the markers discussed was believed to be universally applicable in SLE trials. This section will attempt to summarize some of the proposed biomarkers, but it is not meant to be a complete presentation of all potential markers that can be used in lupus trials. There is no preference for the use of any specific markers, and no specific recommendations can be made. Examples of candidate markers are provided in Table 1 (not all examples are discussed below).

Table 1. Candidate markers in systemic lupus erythematosus
Antibody and autoantibody levels (in serum or tissue)
B cell, T cell, and other cell surface markers
Soluble adhesion molecules and other secreted products
Cytokines and cytokine receptors (in serum or urine)
Complement and breakdown products (in one only serum or tissue)
Genomics (identification of genotypes, Fc and other receptors)
Intracellular protein levels and markers of cellular activation (phosphorylation of proteins/proteomics)
Intracellular RNA (transcriptomics)
Functional assays (lymphocyte function)
Histologic changes on biopsy (such as renal biopsy)

Surrogate markers.

Surrogate markers for clinical outcomes such as mortality and end-stage renal disease were discussed, including measures such as serum creatinine, proteinuria, and hematuria. It appeared that renal histopathology (renal tubular atrophy and interstitial fibrosis, as opposed to the level of inflammatory change) was considered by at least some of the conference participants to be a useful surrogate marker for clinical outcomes. A gold standard for clinical outcomes was needed, and the use of repeat biopsies was discussed as one possible standard. The ethics of re-biopsying patients with clinical remission of their disease (and who have no evidence of active disease) was also discussed, but no clear recommendations were developed.

Examples of biomarkers.

There was a general discussion of autoantibodies as biomarkers in SLE (Dr. D. Pisetsky, Durham, NC) as well as a discussion of one of the oldest biomarkers used in lupus trials, the anti-dsDNA antibody (Dr. M. Petri, Baltimore, MD; Dr. M. Linnik, San Diego, CA). Some autoantibodies are sufficiently SLE specific to be used in classification criteria (anti-dsDNA, anti-Sm, and antiphospholipid antibodies), and subsets of anti-dsDNA antibodies are possibly pathogenic, either through their inclusion in pathogenic circulating immune complexes or in immune complexes formed in situ, or by their damaging of cells either by surface binding or by entry into the cell. In fact, levels of anti-dsDNA antibodies have been used in some clinical studies as markers of disease activity, and they are included in some criteria for activity, such as the SLEDAI. In some studies, increases in anti-dsDNA antibodies appear to correlate with renal flares, and patients with sustained reductions in anti-dsDNA antibodies may have fewer renal flares. However, there is disagreement regarding the sensitivity, specificity, and positive and negative predictive values of measuring anti-dsDNA antibodies.

Similarly, measuring circulating immune complexes has proven to have low specificity and sensitivity for SLE and disease activity. It was suggested that measuring circulating immune complexes containing DNA might provide more clinically useful information. In summary, the measurement of DNA-containing immune complexes in blood is a promising new tool that may prove to be a useful biomarker for disease activity and for response to therapy.

In one study (Dr. J. Buyon, New York, NY), increasing anti-dsDNA antibody titers along with a rise in C3a levels was used to predict clinical flares. It was suggested that individuals with these characteristics can be enrolled in randomized controlled trials and reduction in flare frequency used as an end point. However, it was pointed out that only a very small percentage of patients with increasing titers of these markers will eventually develop a clinical flare. Measurement of anti-dsDNA antibody titers in conjunction with C3a levels may be useful in identifying a subset of patients with a greater likelihood of clinical flare within a relatively short time frame, and may thereby allow shorter trials with greater power to identify efficacy of a new drug. Antibody assays need to be standardized and their cutoff points determined in order to improve the sensitivity and specificity of these assays. These assays should be incorporated into all future clinical trials in lupus.

There was additional discussion about the potential uses of various specific autoantibody determinations. For example, antibodies that bind dsDNA and that also cross-react with N-methyl-D-aspartic acid receptors were identified (Dr. B. Diamond, New York, NY), and it was proposed that these antibodies might contribute to the clinical manifestations of neuropsychiatric lupus. These antibodies have been identified in the central nervous system (CNS) of patients with SLE, and they mediate apoptotic death of neurons in vivo and in vitro (11). However, it appears that a breakdown in the blood–brain barrier is required for the antibody to move into the CNS to induce these changes. Further studies are needed to identify the role of specific antibody subsets and the relevance of their measurement in clinical trials.

The use of cell surface markers was discussed (Dr. P. Lipsky, Bethesda, MD). Antibodies to CD19, CD20, and CD27 were used to divide peripheral B cells into naive (CD19+,CD20+,CD27low or CD27−), memory (CD19+,CD20+,CD27intermediate), and plasma (CD19+,CD20−,CD27high) cells. In this study, lupus patients with flares were noted to have an increase in the percentage of plasma cells in the peripheral B cell population and a reduction in the percentage of naive cells (12). There was a strong correlation between the presence of increased numbers of plasma cells and SLEDAI scores and anti-dsDNA antibody levels. With immunosuppressive treatments, there was a decrease in the plasma cell population and an increase in the memory cell population. Anti-CD154 treatment also led to a decrease in the CD27+ population. B cell markers were also used in an anti-CD20 trial in lupus (Dr. R. Eisenberg, Philadelphia, PA).

An advantage to using markers such as cell surface markers is that changes can be sampled sequentially in individual patients. Different markers may be appropriate for different therapies or even different individuals. However, freshly isolated cells are the ideal material with which to perform these studies, and they may not always be available. The value of any of these types of markers (B cell, T cell, macrophage, etc.) needs to be established in controlled trials, and therapeutic trials should be planned to include serial measurement of these types of markers. Nevertheless, it was concluded that B cell markers may be useful for the diagnosis of SLE and as a measure of disease activity.

The role of B lymphocyte stimulator (BLyS) and its receptors in SLE was presented (Dr. W. Stohl, Los Angeles, CA). BLyS, also called BAFF, TALL-1, THANK, and zTNF4, is a transmembrane protein on myeloid-lineage cells which is released from the cell surface as a biologically active protein and serves as a potent B cell survival factor (13). There are 3 known BLyS receptors, TACI, BCMA, and BAFFR, the last of which is a key receptor for agonist effects on B cells. Theoretically, interfering with the ability of BLyS to signal B cells for survival and Ig secretion could reduce autoantibody levels and be useful in the management of autoantibody-mediated diseases such as SLE. Serum levels of BLyS are elevated persistently or intermittently in approximately one-half of SLE patients. However, those levels do not correlate with disease activity (measured by the SLEDAI), and some patients with RA, Sjögren's syndrome, human immunodeficiency virus infection, and non-Hodgkin's lymphomas also have elevated BLyS levels.

To summarize, BLyS measurements in plasma do not appear to be useful at this time as indicators of response or disease activity in patients with SLE, because many patients with active disease have levels in the normal range, and levels do not parallel disease activity. The major question regarding the clinical role of these molecules is whether targeting certain subsets of patients (such as those with high levels of BLyS) for BLyS blockade with soluble receptors or antibodies can be associated with clinical improvement.

Dr. G. Tsokos (Bethesda, MD) presented data on the use of cytokines to measure disease activity and response to treatment in patients with SLE. A multitude of cytokines participate in immune dysregulation in SLE, with variation between patients and within each patient studied at different times. A general paradigm of the pathophysiology of lupus is that both CD4+ helper T cells and responding B cells are hyperactivated, and regulatory cells, including certain CD4+ subsets, natural killer cells, and CD8+ T cells, are relatively defective. The majority of studies of plasma levels of cytokines, or of cytokines released in vitro by peripheral blood mononuclear cells (PBMCs) spontaneously or following activation, suggest that production of interferon-α (IFNα), interleukin-6 (IL-6), IL-10, and tumor necrosis factor α (TNFα) is increased, while production of IL-1, IL-2, and transforming growth factor β is decreased.

There has been particular interest in IFNα recently, since an IFN-inducible gene has been identified as a major risk factor in NZB-derived strains of murine lupus (14). IFNα-inducible genes have been shown by several groups to be increased in genomic analysis of SLE patient PBMCs compared with PBMCs in normal controls, and disease is prevented in IFNα−/− MRL/lpr mice. IFNγ in murine lupus drives disease, while inhibition of IFNγ by several methods prevents murine lupus; in human SLE, there may also be increased spontaneous production of IFNγ by PBMCs, although some studies show decreased production after cell activation.

As shown by Solomou and coworkers (15) and others, IL-2 production is decreased after in vitro stimulation of PBMCs in some patients with SLE, but this depends in part on the activity of the disease and on the treatment, and on the type of activating stimulus in still other patients, making it difficult to use measurements of this cytokine for general analysis of disease activity or change over time (16–19). On the other hand, some groups have reported that increases in levels of IL-2 receptor (IL-2R) in peripheral blood lymphocytes of patients with lupus correlate with disease activity (19), as do changes in plasma concentrations of intercellular adhesion molecule 1 (ICAM-1) and IL-10. In an example of tracking cytokine and receptor levels as response measures during an intervention, a single open trial of anti–IL-10 in a small number of SLE patients showed uniform decline of disease activity (measured with the Mexican modification of the SLEDAI [20]) during 6 months of followup after 2–3 weeks of treatment. Improvement was accompanied by significant declines in plasma concentrations of soluble IL-2R (sIL-2R), soluble vascular cell adhesion molecule 1, and IL-1 receptor antagonist; spontaneous in vitro production was reduced for sIL-2R and IL-10 (21).

All studies to date measuring changes in cytokines and cytokine receptors in patients with SLE are limited by relatively small numbers and open-trial designs. Appropriate research is required to determine whether these studies add to information from more classic markers of disease activity, such as erythrocyte sedimentation rates, cytopenias, anti-dsDNA antibody titers, and complement levels. However, it is possible that relatively simple measurements of serum levels of IL-2R, ICAM-1, IL-6, IL-10, and TNFα (and possibly IFNα and IFNγ or selected genes they induce) may be satisfactory markers of disease activity and response; appropriate trials to address this possibility are needed.

There was a presentation on the use of erythrocyte C4d and CR1 (E-C4d and E-CR1, respectively) levels as biomarkers in SLE, by Drs. J. Ahearn and S. Manzi (Pittsburgh, PA). When the classical component of complement is activated, C4 is converted to C4a and C4b, which in turn drive activation of C2 with generation of C2b. C4d is a 45-kd molecule activation product from C4 which binds to erythrocytes. During periods of complement activation, quantities of C4d on circulating red blood cells (RBCs) increase. In contrast, erythrocyte surface levels of the CR1 complement receptor decrease as the erythrocytes become saturated with immune complexes. Using a detection assay based on identification of erythrocyte-bound C4d and CR1 with fluorescence-tagged antibodies, it was demonstrated that compared with healthy controls, patients with SLE have reduced proportions of E-CR1 (11% versus 30%) and increased proportions of E-C4d (62% versus 8%). The ratio of E-C4d to E-CR1 is thus higher in SLE patients than in normal controls. These data indicate a high likelihood that the test is useful in distinguishing SLE from other conditions.

Assessing the utility of the assay with regard to SLE disease activity (measured by the SLAM, SLEDAI, and patient's global assessment) by univariate analysis shows a strong correlation of disease activity with E-CR1 levels (P = 0.007) in the absence of statistically significant correlations with E-C4 or with serum levels of C3, C4, and anti-dsDNA antibodies. By multivariate analysis, E-CR1 and anti-dsDNA antibodies were significantly correlated with disease activity. Detection of high proportions of E-C4d may indicate prior disease activity, although it is probably not useful as a pre–disease flare marker. In summary, E-CR1 looks promising as an early marker of disease activity, and reticulocyte-C4d may also be useful in this regard. Since the assay is usually performed on freshly obtained RBCs, there may be a practical limitation for use in multicenter trials.

There was a discussion of the use of complement measurements in trials (Dr. V. M. Holers, Denver, CO; Dr. J. Salmon, New York, NY). Although serum levels of C3 and C4 have been used as markers of active disease in clinical studies, these levels have only 50% sensitivity and 75% specificity for disease flares and do not correlate well with SLEDAI or patient global assessment measures of disease activity. Other measures have been studied, such as C5b–9 levels (the final pathway group of complement split products leading to membrane damage), which correlated with renal flare in some studies. Additional measures have been studied, such as levels of CH50 (normalization of which has been associated with improved renal outcomes), and antibodies to C1q, which occur primarily in patients with proliferative forms of lupus nephritis and which tend to rise when renal flare is occurring. Several problems confound measures of complement or its activation products, including the short half-life of the split products, the role of synthesis of C3 in the liver at the same time it is being consumed by immune complexes, and the response of complement to acute-phase reactants such as IL-6. Since C3 activation is the common target of all the pathways of complement activation (classical, lectin, and alternative), it is reasonable to measure complement components and split products that occur after C3 activation.

In summary, several measurements of plasma levels of complement, antibodies to complement, and split products have been useful in clinical assessments of SLE, particularly in patients with nephritis. However, none of them is sensitive or specific enough at this point to recommend it as a measure of response in clinical trials.

Another approach discussed was the use of microarray analyses of peripheral blood gene expression in SLE patients (Dr. M. Crow, New York, NY; Dr. E. Baechler, Minneapolis, MN). A panel of genes could be used to identify genes that are up-regulated as well as those that are down-regulated in patients with active SLE. Assays are usually performed on fresh cells so as to maintain the integrity of the RNA, which may present some practical issues. These analyses can identify genes that are differentially expressed in clinical subsets of SLE patients, provide insights into the pathogenetic mechanisms of disease, and identify candidate therapeutic targets. For example, genes involved in the IFNα pathway were noted to be up-regulated in patients with active SLE (the IFN-high group is enriched for severe disease) (22), genes involved in IL-8 expression were noted in SLE patients undergoing a clinical flare, and genes such as thrombospondin were identified in patients who appeared to be developing evidence of accelerated atherosclerosis. These approaches may identify subsets of patients who will benefit from therapies directed at cytokine pathways specifically, and from therapies directed at specific mechanisms of disease activity in general.

In summary, the presentations focused on the use of both old and new biomarkers. Clearly, further study is needed before these biomarkers can be used with confidence in lupus clinical trials.

Difficulties with the use of biomarkers/surrogates in lupus trials

  1. Top of page
  2. Introduction
  3. Difficulty studying lupus
  4. Trial designs in lupus
  5. Definitions of surrogate marker and biomarker
  6. Potential roles and uses of biomarkers
  7. Examples of potential surrogates and biomarkers
  8. Difficulties with the use of biomarkers/surrogates in lupus trials
  9. Future approaches
  10. Conclusion
  11. REFERENCES

There are a number of potential sources of variability and differences in biomarkers. For example, biomarker results may vary depending on the patient population examined (e.g., patients selected for specific organ involvement or “all comers”). Biomarkers may vary depending on whether one examines untreated or treated patients or patients with early or late disease. They may also vary with patient ethnicity or with the choice of controls. Two major problems that need to be overcome are the choice of methods of measurement and the lack of standardization of most measurements. Even the measurement of anti-dsDNA antibodies, which has a long history, has not been standardized. Appropriate cutoff points will need to be determined to provide the greatest specificity and sensitivity of these tests.

There may be some further difficulties with the use of biomarkers or surrogate markers in lupus clinical trials. At this time, it is not known whether investigators will need to identify individual biomarkers for each organ studied. In a manner analogous to the case for surrogate markers, a biomarker identified as useful in the study of one drug may not be the same biomarker that is useful for the study of another drug. Therefore, validation is necessary for the use of each biomarker for specific trials. Ultimately, the process of validation and use of biomarkers is time consuming, and this may hinder their use in efficient drug development.

Future approaches

  1. Top of page
  2. Introduction
  3. Difficulty studying lupus
  4. Trial designs in lupus
  5. Definitions of surrogate marker and biomarker
  6. Potential roles and uses of biomarkers
  7. Examples of potential surrogates and biomarkers
  8. Difficulties with the use of biomarkers/surrogates in lupus trials
  9. Future approaches
  10. Conclusion
  11. REFERENCES

There may be strategies to try to address some of the difficulties in the performance of lupus trials raised during this meeting. For example, research consortiums will certainly help in recruiting sufficient numbers of patients and will allow for more efficient use of resources. The NIH has identified 6 areas of high priority for future study: biomarker development, bioinformatics, clinical studies/infrastructure, epidemiology, etiology, and education. This support should certainly help foster development in each of these areas.

A second strategy relates to standardization of tests. Consortiums of investigators will help with the development and use of standardized assays. Once these assays of specific biomarkers are agreed upon and standardized, they should be included in all future trials to allow for validation based on efficacy and safety considerations. Use of the same agreed-upon outcome measures and assays in all future trials in lupus will also allow for greater consistency in trials and for comparison of results between trials. It was proposed that a more immediate alternative approach would be to examine materials (serum, DNA, etc.) from previously performed trials that had incorporated rigorous clinical outcomes to determine whether any markers had correlated with clinical outcomes. However, at present this approach may be problematic in light of the new Health Insurance Portability and Accountability Act requirements for privacy of health information.

Conclusion

  1. Top of page
  2. Introduction
  3. Difficulty studying lupus
  4. Trial designs in lupus
  5. Definitions of surrogate marker and biomarker
  6. Potential roles and uses of biomarkers
  7. Examples of potential surrogates and biomarkers
  8. Difficulties with the use of biomarkers/surrogates in lupus trials
  9. Future approaches
  10. Conclusion
  11. REFERENCES

Although it is difficult at this time to recommend specific markers to include in all clinical trials, the development of surrogate markers and biomarkers is an important goal, especially in the development of new therapies for a disorder as diverse and complex and challenging as SLE. It is hoped that a meeting such as this will spur investigators studying lupus to develop and utilize markers that will lead to improved efficiency of the drug development process. Investigators should be encouraged to include biomarkers appropriate for the end points studied in all future clinical trials in lupus, to allow for validation of these markers as predictors of clinical outcomes. The use of markers in future clinical trials will undoubtedly provide information that will ultimately demonstrate their utility in the drug development process.

REFERENCES

  1. Top of page
  2. Introduction
  3. Difficulty studying lupus
  4. Trial designs in lupus
  5. Definitions of surrogate marker and biomarker
  6. Potential roles and uses of biomarkers
  7. Examples of potential surrogates and biomarkers
  8. Difficulties with the use of biomarkers/surrogates in lupus trials
  9. Future approaches
  10. Conclusion
  11. REFERENCES
  • 1
    Gourley MF, Austin HA 3rd, Scott D, Yarboro CH, Vaughan EM, Muir J, et al. Methylprednisolone and cyclophosphamide, alone or in combination, in patients with lupus nephritis. Ann Intern Med 1996; 125: 54957.
  • 2
    Bansal VK, Beto JA. Treatment of lupus nephritis: a meta-analysis of clinical trials. Am J Kidney Dis 1997; 29: 1939.
  • 3
    Bombardier C, Gladman DD, Urowitz MB, Caron D, Chang CH, and the Committee on Prognosis Studies in SLE. Derivation of the SLEDAI: a disease activity index for lupus patients. Arthritis Rheum 1992; 35: 63040.
  • 4
    Liang MH, Socher SA, Larson MG, Schur PH. Reliability and validity of six systems for the clinical assessment of disease activity in systemic lupus erythematosus. Arthritis Rheum 1989; 32: 110718.
  • 5
    Symmons DP, Coppock JS, Bacon PA, Bresnihan B, Isenberg DA, Maddison P, et al, and Members of the British Isles Lupus Assessment Group (BILAG). Development and assessment of a computerized index of clinical disease activity in systemic lupus erythematosus. QJM 1988; 69: 92737.
  • 6
    Vitali C, Bencivelli W, Isenberg DA, Smolen JS, Snaith ML, Sciuto M, et al, and The European Consensus Study Group for Disease Activity in SLE. Disease activity in systemic lupus erythematosus: report of the Consensus Study Group of the European Workshop for Rheumatology Research. II. Identification of the variables indicative of disease activity and their use in the development of an activity score. Clin Exp Rheumatol 1992; 10: 5417.
  • 7
    Urowitz MB, Gladman DD, Tozman EC, Goldsmith CH. The lupus activity criteria count (LACC). J Rheumatol 1984; 11: 7837.
  • 8
    Strand V, Gladman D, Isenberg D, Petri M, Smolen J, Tugwell P. Outcome measures to be used in clinical trials in SLE. J Rheumatol 1999; 26: 4907.
  • 9
    Gladman D, Ginzler E, Goldsmith C, Fortin P, Liang M, Urowitz M, et al. The development and initial validation of the Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index for systemic lupus erythematosus. Arthritis Rheum 1996; 39: 3639.
  • 10
    Ware JE Jr, Snow KK, Kosinski M, Gandek B. SF-36 health survey: manual and interpretation guide. Boston: The Health Institute, New England Medical Center; 1993.
  • 11
    DeGiorgio LA, Konstantinov KN, Lee SC, Hardin JA, Volpe BT, Diamond B. A subset of lupus anti-DNA antibodies cross reacts with the NR2 glutamate receptor in systemic lupus erythematosus. Nat Med 2001; 7: 118993.
  • 12
    Jacobi AM, Odendahl M, Reiter K, Bruns A, Burmester GR, Radbruch A, et al. Correlation between circulating CD27high plasma cells and disease activity in patients with systemic lupus erythematosus. Arthritis Rheum 2003; 48: 133242.
  • 13
    Stohl W. SLEsystemic lupus erythematosus: a BlySful yet BAFFling disorder. Arthritis Res 2003; 5: 1368.
  • 14
    Rozzo SJ, Allard JD, Choubey D, Vyse TJ, Izui S, Peltz G, et al. Evidence for an interferon-inducible gene, Ifi202, in the susceptibility to systemic lupus. Immunity 2001; 3: 43543.
  • 15
    Solomou EE, Juang YT, Gourley MF, Kammer GM, Tsokos GC. Molecular basis of deficient IL-2 production in T cells from patients with systemic lupus erythematosus. J Immunol 2001; 166: 421622.
  • 16
    Alcocer-Varela J, Alarcon-Segovia D. Longitudinal study on the production of and cellular response to interleukin-2 in patients with systemic lupus erythematosus. Rheumatol Int 1995; 15: 5763.
  • 17
    Murakawa Y, Sakane T. Deficient phytohemagglutinin-induced interleukin-2 activity in patients with inactive systemic lupus erythematosus is correctable by the addition of phorbol myristate acetate. Arthritis Rheum 1988; 31: 82633.
  • 18
    Draeger AM, Swaak AJ, van den Brink HG, Aarden LA. T cell function in systemic lupus erythematosus: normal production of and responsiveness to interleukin 2. Clin Exp Immunol 1986; 64: 807.
  • 19
    Wigfall DR, Sakai RS, Wallace DJ, Jordan SC. Interleukin-2 receptor expression in peripheral blood lymphocytes from systemic lupus erythematosus patients: relationship to clinical activity. Clin Immunol Immunopathol 1988; 47: 35462.
  • 20
    Guzman J, Cardiel MH, Arce-Salinas A, Sanchez-Guerrero J, Alarcon-Segovia D. Measurement of disease activity in systemic lupus erythematosus: prospective validation of 3 clinical indices. J Rheumatol 1992; 19: 15518.
  • 21
    Llorente L, Richaud-Patin Y, Garcia-Padilla C, Claret E, Jakez-Ocampo J, Cardiel MH, et al. Clinical and biologic effects of anti–interleukin-10 antibody administration in systemic lupus erythematosus. Arthritis Rheum 2000; 43: 1790800.
  • 22
    Baechler EC, Batliwalla FM, Karypis G, Gaffney PM, Ortmann WA, Espe KJ, et al. Interferon inducible gene expression signature in peripheral blood cells of patients with severe lupus. Proc Natl Acad Sci U S A 2003; 100: 26105.