Diagnosis of Lupus

Introduction

Timely diagnosis of systemic lupus erythematosus (SLE) is of critical clinical importance because it allows for an early start to treatment, preventing organ damage and improving the patient’s quality of life. Unfortunately, the spectrum of clinical manifestations of SLE is so broad, and the individual patterns of disease so diverse – changing with time even in individual patients – that diagnosis is often difficult to obtain. Because diagnosis requires a high degree of clinical expertise to combine the clinical and immunological findings, it is often delayed, taking in some cases 6 or more years. Since no definitive diagnostic criteria for SLE exist, physicians often rely on classification criteria – developed for use in scientific research – as a guide to diagnosis. While useful, overreliance on or misinterpretation of the classification criteria can lead to diagnostic errors. In this module, we present the most widely used classification criteria and give a brief overview of the diagnostic process.

SLE Classification Criteria and Diagnosis

Classification is a concept employed for research purposes – using a definition based on a relatively limited number of criteria to find relatively uniform group of patients that can be used to scientifically compare treatments. Diagnosis is a highly iterative process of excluding other entities (differential diagnoses) and using all of the available information to provisionally categorize one patient. Specificity is more important in classification because excluding true negative cases from research populations improves the validity of the scientific findings. Sensitivity is more important in diagnosis, because identifying true positive cases allows those patients to receive proper care. Although classification and diagnosis are distinct concepts, classification criteria nevertheless provide useful information in the diagnostic process, if properly understood and applied. Conversely, if misused in diagnosis, classification criteria can hinder proper care. For example, healthcare providers who are not rheumatologists may consider the mere presence of anti-dsDNA as diagnostic for SLE, despite the fact that such antibodies are relatively common in the general population. Because of its complex clinical presentation, SLE is a difficult disorder to study scientifically in the context of clinical trials; recruiting a homogeneous population of patients can be challenging. To help the clinical trial selection and inclusion process, several classification criteria for SLE have been developed. The first of these are the American College of Rheumatology (ACR) criteria, first proposed in 1982 and then updated in 1997. The 1997 ACR criteria have been, and continue to be, widely used globally in clinical trials and epidemiological studies. To be classified as having SLE, a patient must fulfill 4 out of the 11 criteria (Table 5-1). In 2012 the Systemic Lupus International Collaborating Clinics (SLICC) developed their own set of criteria to improve some of the shortcomings of the 1997 ACR criteria, including duplication of cutaneous features, omission of certain neurologic manifestations and immunologic abnormalities, and – most concerningly – the possibility of a patient with renal biopsy-confirmed lupus nephritis (LN) to nevertheless fail to classify as having SLE. The SLICC classification requires a patient to fulfill 4 out of 17 criteria, with at least 1 of the 11 clinical and 1 of the 6 immunologic criteria, or to have biopsy-proved nephritis in the presence of anti-nuclear autoantibodies (ANAs) or double-stranded DNA-directed (anti-dsDNA) antibodies (Table 5-2). The most recent set of criteria were developed by the European League Against Rheumatism (EULAR) in collaboration with the ACR, and published in 2019. The main goal of the 2019 EULAR/ACR criteria was to improve the sensitivity (the probability of detecting a true case of SLE) and specificity (the probability of excluding a case that really is not SLE) of earlier classification criteria and to improve detection of SLE in the early stages of the disease. Applied to a validation cohort, the 2019 EULAR/ACR criteria had a sensitivity of 96.1% and specificity of 93.4%, while the 1997 ACR criteria had 82.8% sensitivity and 93.4% and the SLICC criteria had 96.7% sensitivity and 83.7% specificity. In the 2019 EULAR/ACR classification system, the presence of an ANA titer of at least 1:80 is counted as an essential “entry criterion” (see Testing for SLE-associated antibodies and Terminology of SLE-associated autoantibodies in Pathogenesis and Pathophysiology for a more detailed discussion on ANAs and other SLE-associated antibodies). Other, “additive” criteria are grouped into 7 clinical and 3 immunologic groups and each criterion is assigned a point weight of 2-10 points. A classification of SLE is given if a patient satisfies the entry criterion and has a total point score of 10 or higher in the additive criteria (Figure 5-1).

Enlarge 

Enlarge 

Enlarge 

The Diagnostic Process

The diagnosis of SLE is made challenging by the heterogeneous spectrum of clinical manifestations, multi-organ involvement, and unpredictable disease course with intermittent signs and symptoms. This often necessitates a long observation period before a definitive diagnosis can be made. A diagnosis of SLE typically includes integrating both clinical and laboratory findings and excluding differential diagnoses. A proposed diagnostic algorithm, geared towards use in the primary care setting, is shown in Figure 5-2.

The process begins by examining the clinical features and the demographics of the patient; the latter is important given the predominance of women of childbearing age among persons with SLE, and the higher prevalence of the disease in non-White populations (see the Epidemiology section). It is important to remember that SLE is a multi-organ disorder, and may present with a variety of symptoms, including constitutional, cutaneous, renal, neuropsychiatric, hematologic, cardiovascular, pulmonary and others, many of which are non-specific. The most common symptoms are constitutional and non-specific, such as fatigue, weight loss, fever without infection and arthralgia. Other potentially significant symptoms include malar rash (31%), photosensitivity with (sub)acute cutaneous rash (23%), pleuritic chest pain (16%), Raynaud phenomenon – decreased blood flow to the fingers – (16%) and mouth sores (12.5%). History taking and physical examination will reveal many of these features and form a basis for suspected SLE. A number of rheumatic, infectious and inflammatory diseases have symptoms that overlap with SLE, and it is important to exclude these other causes while pursuing a potential diagnosis of SLE. Common differential diagnoses are shown in Table 5-3, along with their distinguishing features and diagnostic modalities.

Once SLE is suspected on the basis of clinical presentation, it is reasonable to obtain an ANA titer; while it is possible for SLE to be diagnosed without ANAs, this situation is exceedingly rare. If the patient is positive for ANAs, further laboratory testing is indicated, including complete blood count, urinalysis and tests for specific-lupus associated antibodies. Antibodies with good diagnostic specificity for SLE include anti-dsDNA and anti-Smith (anti-Sm) antibodies, but other antibodies may be useful as well, including anti-histone (associated with drug-induced lupus), anti-Ro/anti-La (associated with neonatal lupus), anti-ribosomal P (associated with neuropsychiatric manifestations), and anti-phospholipid antibodies (associated with thrombosis, recurrent miscarriages and neurologic disease). Approximately 50% of patients with SLE also show high type I IFN expression, which may be associated with high disease activity. Thus, IFN represents another, albeit non-specific, biomarker that could be used in the diagnostic process; however, no commercial products are as yet available. Cell-bound complement activation products (CB-CAPs) have also been investigated as a potential biomarker; patients with SLE show higher flow cytometry-measured CB-CAP levels than patients with other diseases. A CB-CAP test developed by Avise^® is commercially available.

Once the clinical and laboratory findings have been collected and are consistent with a provisional diagnosis of SLE, this information can be used to apply SLE classification criteria, such as the 2019 EULAR/ACR criteria (Figure 5-2). It is important to stress once more that while classification criteria are a good diagnostic aid, they were designed for a research purpose, and should therefore only guide diagnostic thinking. Many features that may be diagnostically useful are not queried by the criteria, including, for example, arthralgia, lymphopenia, peripheral neuropathy, or the presence of anti-ribosomal P antibodies; see Table 5-4 for a list of features other than those listed in the 2019 EULAR/ACR criteria that are still useful when considering a diagnosis of SLE. Finally, it should be re-emphasized that no feature has 100% diagnostic specificity. Even the absence of ANAs, which are absolutely required in the 2019 EULAR/ACR criteria, does not definitively exclude SLE, although it renders it very unlikely.

Enlarge 

Enlarge 

Enlarge