INTRODUCTION

Systemic lupus erythematosus has a characteristically waxing and waning course that can range in severity from a relatively mild chronic condition to an acutely life-threatening disease. The clinical heterogeneity most likely reflects the combination of various pathogenetic events modified by the patient’s genetic background, environmental factors, and treatment. This complexity adds to the difficulty of diagnosing and optimally managing patients. Biologic markers that reliably describe various aspects of SLE could be used to improve diagnostic accuracy, predict prognosis, characterize disease activity and guide therapy. Markers that are closely associated with or predict important clinical outcomes could be used as surrogate endpoints in clinical trials to accelerate drug development. Numerous candidates have been proposed as biomarkers, primarily based on data reflecting various pathophysiologic abnormalities; however, there are no biomarkers that have been rigorously validated and are widely accepted in SLE. The need to use biomarkers as surrogate endpoints in clinical trials is becoming more pressing with the availability of several promising new therapeutic agents and the unwillingness of the pharmaceutical industry to embark on studies that may last 5 to 10 years to show a beneficial outcome using traditional clinical endpoints. In this chapter, we will provide a general overview of biomarkers and surrogate endpoints, and summarize our view on the most promising candidates of biomarkers in SLE with a focus on promising candidate biomarkers of disease activity.

BIOMARKERS AND SURROGATE ENDPOINTS

The terms biomarker and surrogate endpoint describe different entities. However, they are commonly used interchangeably, which has led to considerable confusion in the literature. In an attempt to prevent such confusion and standardize the nomenclature, the National Institutes of Health convened an expert panel in 1991. Their recommendations are summarized in Table 5.1.¹ Biomarker can be defined as a physical sign or cellular, biochemical, molecular, or genetic alteration by which a normal or abnormal biologic process can be recognized and/or monitored, and that may have diagnostic or prognostic utility. Biomarkers must measure an underlying biologic process reliably and reproducibly. Surrogate endpoint is a measurement that is intended to serve as a substitute for a clinically meaningful outcome and is expected to predict the effect of a therapeutic intervention. Both biomarkers and surrogate endpoints have to be validated to prove that they are measuring intended outcomes reliably. It is important to recognize that the requirements of surrogate markers are much more stringent and that only a small minority of biomarkers will fulfill the criteria of a surrogate endpoint. In order for a biomarker to be validated as a surrogate endpoint, it must be shown that the presence of or a change in the measurement predicts an important clinical endpoint.

TABLE 5.1 VALIDATION OF BIOMARKERS AND SURROGATE ENDPOINTS

Term	Definition¹	Elements of Validation
Biomarker	A characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention.	Sensitivity: the level of correlation between the magnitude of change in the biomarker and clinical/biological endpoint Specificity: the extent to which a biomarker explains the changes in a clinical/biological endpoint Bioanalytical assessment: the laboratory test or measurement should include assessment of precision, reproducibility, range of use, variability, and practicality False positivity: a change in a biomarker is not reflected by a change in a clinical/biological endpoint False negativity: a biomarker does not change despite a change in the clinical/biological outcome
Surrogate endpoint	A surrogate endpoint is expected to predict clinical benefit or harm (or lack of benefit or harm) based on epidemiologic, therapeutic, pathophysiologic, or other scientific evidence. A biomarker may fulfill the criteria for a surrogate endpoint.	Requires evaluation and qualification ^1,²: a graded process by which evidence is acquired linking a biomarker with a clinical endpoint Biologic plausibility: a mechanistic basis for using a surrogate endpoint Statistical relationship between the biomarker and the clinical endpoint based on epidemiologic or observational studies of the natural history of the disease Estimate of the expected benefit based on adequate, well-controlled and appropriately powered clinical trials Ability to predict potential adverse reactions Consistency of effects following interventions with various drug classes and/or within different stages of disease

VALIDATION OF BIOMARKERS

Validation of biomarkers is a complex process.^2,³ The criteria for validation should be defined by the nature of the question that the biomarker is intended to address, the degree of certainty required for the answer, and the assumptions between the biomarker and clinical endpoints. An ideal biomarker should measure a clinically relevant process, and be sensitive and specific for the measurement that it is intended for.

Any biomarker should be validated for sensitivity, specificity, details of bioanalytical assessment, and the probability of false positives and false negatives. Sensitivity is the ability of a biomarker to reflect a meaningful change in important clinical and/or biological endpoints and describes the level of correlation between the magnitude of change in the biomarker and clinical/biological endpoint. However, even a strong correlation does not prove a cause–effect relationship. Specificity defines the extent to which a biomarker explains the changes in a clinical/biological endpoint. The bioanalytical assessment of the laboratory test or measurement should include assessment of precision, reproducibility, range of use, variability, and practicality. False positivity is the situation in which a desired change in a biomarker is not reflected by a positive change in a clinical/biological endpoint, or even worse, is associated with a negative change. False negativity is the opposite, when a biomarker does not change despite a change in the clinical/biological outcome (Table 5.1).

Qualification of Surrogate Endpoints

An ideal surrogate endpoint can be thought of as a validated biomarker that can be definitively substituted for a clinically meaningful endpoint in an efficacy trial or clinical practice. To meet the most rigorous standards, the surrogate endpoint must correlate with the true clinical outcome and must fully capture the net clinical effect of treatment.⁴ This may be impossible to achieve for most biomarkers, but it is clear that extensive clinical evidence is needed, which is collected in a rigorous scientific process and analyzed by careful statistical assessment. There is no consensus on validation of surrogate endpoints, and some experts favor the term qualification to describe this process,^1,⁵ which has to include the following elements. Biologic plausibility should provide a mechanistic basis for using a surrogate endpoint, and epidemiologic or observational studies of the natural history of the disease should establish the statistical relationship between the biomarker and the clinical endpoint. Adequate, well-controlled and appropriately powered clinical trials should provide an estimate of the expected benefit; ideally, an appropriate dose– or exposure–response relationship would provide additional support for surrogate status. The analysis should include a consideration of whether potential adverse reactions are predicted by the surrogate endpoint. It is essential that the development and validation of biomarkers and surrogate markers be built into the drug development process, starting from the preclinical phase. It may be helpful to conduct a meta-analysis of multiple clinical trials to determine the consistency of effects following interventions with various drug classes and within different stages of disease.²

Current State of Biomarkers in SLE

A large number of studies described potential biomarkers in lupus, but none fulfills the criteria of a true validated biomarker. There are many reasons that account for the conflicting results of various studies. Most were not designed as biomarker validation studies; therefore, their study design may not be appropriate for this purpose. A cross-sectional study may show an association between a biomarker and disease activity in a group of patients, but longitudinal studies are necessary to evaluate whether the same marker can be used to monitor disease activity in individual patients. The patient population may differ among studies in several ways such as ethnicity, treatment, organ manifestation, and stage of disease (early vs. late). The choice of controls (healthy vs. other rheumatic diseases vs. subpopulations of lupus patients) is also critical, and varies substantially among studies. Most frequently, widely accepted disease activity indices, such as SLEDAI, SLAM, ECLAM, and BILAG, are used as outcome measures. Although all are valid tools, they do not capture exactly the same aspects of lupus.⁶ Therefore, in some studies selected biomarkers correlated with one but not another activity index. In some cases, organ-specific outcome measures may be more appropriate, but the lack of widely accepted clinical endpoints makes standardization difficult. A large number of studies lack the statistical rigor that is essential to draw valid conclusions. Last, but not least, the bioassays used to measure biomarkers are frequently not standardized, leading to conflicting results in different laboratories.

TABLE 5.3 BIOMARKERS OF SLE DISEASE ACTIVITY

BIOMARKERS IN SLE

Lupus is a chronic disease with various mechanisms dominating at various stages. Genetic predisposition and environmental factors are the most important for predicting or quantifying the risk of lupus. At the onset of clinical symptoms, classic biomarkers of autoreactivity such as autoantibodies are used to help to establish or confirm the diagnosis of SLE. During the acute phases of lupus, biomarkers of disease activity could be used to optimize risk–benefit assessment in individual patients. Eventually, the sustained autoimmune and inflammatory process leads to the next phase characterized by organ damage, when there is a much greater reliance on physiologic measures of organ function, such as renal function measurements.

Due to the complexity of the underlying pathogenesis, it is likely that any specific biomarker will perform best at certain stages of the disease and novel biomarkers have to be defined specifically as to what they are intended to reflect (prognosis, future organ involvement, severity, disease activity, risk of flare-up, etc.), and at what stage of the disease (Table 5.2).

TABLE 5.2 POTENTIAL USES OF BIOMARKERS IN SLE

Application	Potential Biomarkers
Predict SLE and/or organ involvement	Genetic markers Autoantibodies
Monitor disease activity	Soluble markers (blood, urine, cerebrospinal fluid) Autoantibodies Complement activation markers Cytokines Chemokines Mediators of inflammation Circulating subsets of inflammatory cells Effector T and B lymphocytes Regulatory T cells Complex gene expression or proteomic profiles
Predict response to therapy^a	Genetic markers Related to lupus Related to drug metabolism (pharmacogenomics) Changes in selected markers of diseases activity
Predict flare^a	Changes in selected markers of diseases activity
Predict damage^a	Genetic markers Autoantibodies Complex gene expression or proteomic profiles Lack of normalization of markers of disease activity
Describe damage	Physiologic measures Imaging

a Therapeutic decision based on predictive biomarkers should be made only after a strong correlation between the biomarker and the clinical outcome has been established; that is, the biomarker is qualified as a surrogate endpoint.

A detailed review of the literature on novel biomarkers in SLE was published recently.^7,⁸ Other chapters in this book describe the genetics of lupus and the current use of various laboratory tests in diagnosing SLE and their relationship to specific clinical manifestations. Here we will provide a summary of both the conventional and the most promising biomarkers of disease activity.

Recent studies have demonstrated that autoantibodies can be found in the majority of lupus patients years before the diagnosis or the onset of symptoms.^9,¹⁰ Moreover, the appearance of antibodies seemed to follow a temporal sequence. In 88 of the 130 patients with SLE, at least one autoantibody tested was present before the diagnosis (up to 9.4 years earlier; mean, 3.3 years). High titer antinuclear antibodies were present in 78%; anti–double-stranded DNA and anti-Ro antibodies in about 50%; anti-La, anti-Sm, and antinuclear ribonucleoprotein in about 30%; and antiphospholipid antibodies in 18% of patients. Antinuclear, antiphospholipid, anti-Ro, and anti-La antibodies were present earlier than anti-Sm and antinuclear ribonucleoprotein antibodies (a mean of 3.4 years before diagnosis vs. 1.2 years). Anti–double-stranded DNA antibodies, with a mean onset 2.2 years before the diagnosis, were found later than antinuclear antibodies (p=0.06) and earlier than antinuclear ribonucleoprotein antibodies. The finding that the appearance of autoantibodies in patients with SLE tends to follow a predictable course, with a progressive accumulation of specific autoantibodies before the onset of SLE raises the possibility of using autoantibody profiles as predictors of lupus. These observations are clearly exciting; however, their clinical applicability needs to be determined.

Although they account only for a very small proportion of patients, deficiencies in early complement components, such as C4, C2, and C1q, are associated with a significantly increased susceptibility to lupus. Moreover, activation of the complement system plays a central role in the pathogenesis of lupus. Therefore, it is conceivable that subtle changes in the complement system could be used for early diagnosis of SLE. A recent study showed that combined detection of high levels of erythrocyte-bound C4d (E-C4d) and low levels of erythrocyte-complement receptor 1 (E-CR1) had a sensitivity of 81% and specificity of 91% in distinguishing SLE patients from normal controls and acceptable sensitivity (72%) and specificity (79%) in distinguishing SLE and other autoimmune diseases. The overall negative predictive value of the combination of the two tests was 92%.¹¹^–¹³ Platelet C4d was found in only 18% of lupus patients, but it had 100% specificity against normal subjects and 98% against patients with other autoimmune diseases.¹⁴ Both of these findings have to be confirmed in larger studies, and their applicability to early diagnosis needs to be formally tested.