INTRODUCTION

The abundant use of classification criteria in rheumatology is a mixed blessing for practicing physicians.¹ In the case of systemic lupus erythematosus, the introduction and subsequent widespread application of the initial, revised and updated American College of Rheumatology (ACR) criteria²^–⁴ has been instrumental in delineating well-defined patient cohorts for epidemiologic, outcome, and intervention studies. However, as clearly stated in relevant publications, these criteria are the result of extensive statistical modeling to reach the lowest amount of heterogeneity in study cohorts.

SLE on the other hand remains a clinical syndrome with a diverse phenotype in each individual patient that is in addition often variable over time. In daily practice, this has (silently) evolved into a tradition, where the ACR criteria are needed to confirm the clinical diagnosis of SLE. This practice has never been validated and has several drawbacks. First, patients not fulfilling four of the required ACR criteria could possibly be not given the diagnosis (and not receive the required treatment). It is of the utmost importance to realize that a large amount of data reduction was applied to achieve the models supporting the ACR criteria. In the process, a lot of additional clinical information has been excluded.² These clinical data may nonetheless prove useful in the management of an individual patient, regardless of the amount of classification criteria fulfilled. Table 43.1 clearly demonstrates the problem: not only may patients present with a multisystemic disease and still not fulfill the ACR criteria, but a patient may also fulfill four criteria while the diagnosis of SLE is quite questionable. For example, the current ACR criteria contain four dermatologic elements, and thus a patient with only skin involvement can be diagnosed with SLE. However, it is well known that only a few (5%) of chronic cutaneous lupus patients will develop a more severe systemic disease. As a partial solution to this, methods have been proposed of weighting the various ACR criteria,^5,⁶ but even this system failed to take into account that various disease presentations do not occur independently of one another. The criteria photosensitivity and malar rash are highly interrelated, while there also is a strong correlation between anti-dsDNA Ab and positive antinuclear antibody (ANA) results.⁷

TABLE 43.1 FREQUENT DISEASE FEATURES IN SLE AND ACR CLASSIFICATION CRITERIA

Organ System	ACR Classification Criteria	Other Frequent Features
Constitutional	—	Fever, malaise, anorexia, weight loss
Cutaneous	1 Malar rash 2 Discoid rash 3 Photosensitivity 4 Oral/nasopharyngeal ulcers	Alopecia, Raynaud’s subacute cutaneous LE, urticaria, bullous lesions, vasculitis, panniculitis,
Musculoskeletal	5 Nonerosive arthritis	Arthralgia, myalgia, myositis, tendonitis, ligamentous laxity, osteonecrosis
Cardiopulmonary	6 Pleuritis/pericarditis	Pleural effusions, myocarditis, pneumonitis, verrucous endocarditis, interstitial fibrosis, pulmonary hypertension, dyspnea
Renal	7 Proteinuria (>500 mg/day) Urinary cellular casts	Nephrotic syndrome, renal insufficiency, renal vein thrombosis, hypertension
Neurologic	8 Psychosis Seizure	Organic brain syndrome, cranial neuropathies, peripheral neuropathies, cerebellar signs
Gastrointestinal	—	Pancreatitis, ascites, elevated liver enzymes, vasculitis
Hematologic	9 Hemolytic anemia Leucopenia (<4000) Lymphopenia (<1500) Thrombocytopenia (<100,000)	Anaemia of chronic disease, lupus anticoagulant, thrombosis, splenomegaly, lymphadenopathy
Other systems	—	Sicca syndrome, conjunctivitis, episcleritis
Laboratory	10 Anti-dsDNA Ab, anti-Sm Ab, False-positive VDRL/ApL Ab 11 Positive ANA	Lupus band test on skin biopsy, anti-C1q Ab, hypocomplementemia

In the past, patients with typical symptoms but not fulfilling the ACR criteria for SLE or another connective tissue disease (CTD), have been labeled with diagnoses such as subclinical lupus, variant lupus, latent lupus, or overlap syndromes, with the overall idea that these patients would develop a specific disease entity in the future.⁸^–¹³ In our opinion, incomplete lupus (ILE) is the most fitting description for these patients’ condition and this chapter summarizes current knowledge on this group of patients. It is our aim to contribute to a uniform understanding of this genuine clinical entity, which is well known to experienced rheumatologists, but has largely been identified in earlier textbooks.¹⁴ Generating further interest and study can facilitate the management of these patients with better and earlier identification of patients at risk of progressing to SLE. Finally, knowledge about the manner in which immune disturbances in ILE patients translate into a less severe clinical phenotype may provide important clues for the whole field.

DELINEATING INCOMPLETE LUPUS ERYTHEMATOSUS

SLE is usually a disease that evolves over time and only a minority of patients present with multisystem disease that can be promptly classified according to the ACR criteria. In the LUMINA study cohort, this occurred in only 15 % of patients.¹⁵ In the remainder of patients, the mean time from the fulfillment of the first ACR criterion to the fulfillment of at least four criteria varied from 39 months for patients with one presenting ACR criterion to 32 and 14 months, respectively, for patients presenting with two or three ACR-validated criteria.

The majority of these SLE patients were thus initially not classified as having SLE, but rather as having ILE for various periods of time, and it would be interesting to know if presenting with ILE had therapeutic consequences. Similar timeframes for ILE have been described in European studies as well.^13,^16,¹⁷ Also of note was the longest observed lag time in the LUMINA cohort of 328 months, illustrating the potential for late disease progression in ILE patients. While these and other data illustrate the gradual progression from ILE to SLE in selected patients, they do not provide much needed information on patients not progressing beyond three ACR criteria. There are limited data in the literature that address this question. A few studies have analyzed the frequency and course of ILE separately, while other studies have included ILE in the group of undifferentiated connective tissue diseases (UCTDs). The term “undifferentiated connective tissue diseases” was introduced in the 1980s as the widespread use of the ANA test identified increasing numbers of patients with a systemic disease and a positive ANA test, and the question was raised whether these patients all really had SLE or that milder disease forms were more often recognized with the widespread diagnostic ANA testing. The further refinement of serologic test systems to demonstrate more subgroups of ANA indicate that patient groups potentially could be characterized by their specific antinuclear antibody as well as titer. However, criteria for the diagnosis of UCTD are not yet established and the name should be reserved for those patients with features strongly suggestive of an autoimmune rheumatic disease but not readily classified or related to a specific CTD.¹⁸ Published UCTD series contain variable numbers of ILE patients, but in addition, groups of patients with not yet fully recognized rheumatoid arthritis or scleroderma.^19,²⁰ The question whether ILE is a separate entity or only a subset of UCTD is difficult to answer due to a lack of data on the subject, and reflects our basic uncertainty how to regard this group of patients in the larger scheme of things (Figure 43.1). Only one study has reported epidemiologic data on ILE.¹³ In a population based in Denmark, the prevalence of ILE was estimated to be 5.2 per 100,000, which was about one-quarter the prevalence of definite SLE I in that region. This rate must be considered a minimum estimate given the selection bias towards definite SLE in the study design. In Sweden, 28 ILE patients were identified during a 10-year observation period; from the numbers given, an annual ILE incidence of almost 2 per 100.000 can be inferred.^12,²¹ Other cross-sectional studies on ILE do not report epidemiologic figures, but conclude that ILE is not infrequent.^9,^10,²² Despite a larger number of studies on UCTD in the literature, these are all cross-sectional in design and give no indication of the relative frequency of this condition.²³^–²⁶