Pre-Clinical Lupus

Abstract

Purpose of review

Systemic lupus erythematosus (SLE) is often preceded by immune dysregulation and clinical manifestations below the threshold for SLE classification. This review discusses current and evolving concepts about the pre-classification period of SLE, including clinical and mechanistic observations, and potential avenues for early identification and intervention.

Recent findings

Although incomplete lupus erythematosus (ILE) involves fewer clinical manifestations than SLE, ILE can cause organ damage and mortality. Common clinical features in ILE include antinuclear antibody seropositivity, polyarthritis, immunologic manifestations, and hematological disorders. Despite having lower disease activity and damage scores than SLE patients, ILE patients may develop pulmonary arterial hypertension or renal, neurological, or peripheral vascular damage. The recently proposed SLICC SLE classification criteria could shift the period considered “preclinical SLE”. Murine studies suggest that the balance of T helper/T regulatory cells, peroxisome proliferator-activated receptor γ activity, and plasmacytoid dendritic cell pathways may be valuable targets for early intervention.

Summary

Advances in our understanding of early SLE, including stages before clinical features are fully developed, will improve our ability to identify individuals at high risk of classification for potential prevention trials, provide necessary information to improve diagnostic testing, and perhaps identify novel targets for directed therapeutics in clinical SLE.

Introduction

Systemic lupus erythematosus (SLE) is a complex disease characterized by autoantibody production and remarkably heterogeneous clinical presentations, with the potential involvement of nearly every organ system. Therefore, to facilitate diagnosis and aid in the identification of SLE patients for clinical studies, SLE is typically classified based on the presence of at least four of eleven clinical or serologic parameters defined by the American College of Rheumatology (ACR) (1–3). While some patients meet ACR classification criteria for SLE at their initial presentation to the healthcare system, others present with a limited number of criteria that are consistent with SLE, yet insufficient for formal classification. Such patients may either transition to classified SLE through the accumulation of additional criteria, or follow a more limited and stable disease course that never meets classification criteria.

Although patients who transition to classified SLE from a clinically apparent pre-SLE period often remain free of nephritis or major CNS involvement (4–10), the majority of SLE patients have irreversible organ damage by the time of diagnosis. It is thought that early therapeutic intervention, before the onset of permanent damage, could delay or prevent the transition to SLE and reduce lupus-associated organ damage. In aggregate, multiple studies demonstrate that select demographic, clinical, and serologic features may distinguish individuals at the highest risk for subsequent SLE development (11, 12), but additional studies are needed to specifically define and validate potential predictors. Though challenging, research into the earliest stages of SLE would also support the development of directed diagnostics and biomarkers for identification of early disease activity before irreversible target-organ damage. Here we review recent advances in our understanding of the clinical features, pathogenetic changes, and potential interventions in pre-classified SLE.

Clinical manifestations in incomplete lupus

Although incomplete lupus erythematosus (ILE) is sometimes considered a mild form of lupus and may be a precursor to complete SLE, the clinical manifestations of ILE can be severe (Table 1). This may explain why many ILE patients are treated with immunomodulatory medications (Table 2). The Systemic Disease Study Group of the Spanish Society of Rheumatology (EAS-SER) recently described the clinical characteristics of incomplete lupus patients in the multi-center Spanish Rheumatology Society SLE Registry (RELESSER) (13). The study included 345 ILE patients, who had fewer than four 1997 ACR-SLE criteria and a clinical diagnosis of SLE by an experienced rheumatologist, as well as 3679 SLE patients, who met at least four ACR-SLE criteria. All criteria were more common in SLE patients than ILE patients. The most common criteria in ILE patients were ANA (94.7% of ILE patients), immunological disorder (55.1%), arthritis (44.2%), and hematological disorder (43.5%). In addition, 17.3% of ILE patients had cardiac manifestations, 14.9% had neuropsychiatric symptoms, and 6.3% had lupus nephritis (13). Disease activity, damage, and severity were lower in ILE patients compared to SLE patients when adjusted for gender, age at diagnosis, and disease duration. However, disease activity led to hospitalization in 28.0% of ILE patients. In addition, 5 of 15 deaths in ILE patients (33.3%) were attributed to SLE activity, with an average age at death of 71.7 years (13).

Table 1.

Percentage of incomplete or potential lupus patients meeting select ACR-SLE criteria.

Cohort	Spanish Rheumatology Society Lupus Registry N=345*	Peking Union Medical College Hospital N=77*	Brigham and Women’s Hospital N=161*
Malar rash, %	11.0	9.1	11
Discoid rash, %	7.1	0	1
Photosensitivity, %	20.5	1.3	22
Oral ulcers, %	7.3	7.8	11
Arthritis, %	44.2	20.8	57
Serositis, %	8.8	20.8	13
Renal disorder, %	4.3	16.9	1
Neurological disorder, %	1.2	6.5	4
Hematological disorder, %	43.5	51.9	30
Immunological disorder, %	55.1	29.9	+
ANA, %	94.7	97.4	98

^*Please see text and references for additional details about the Spanish Rheumatology Society Lupus Registry (13), Peking Union Medical College Hospital (14), and Brigham and Women’s Hospital (15) cohorts.

⁺Reported as separate values for anti-dsDNA (16%), anti-Sm (3%), anticardiolipin (14%), and lupus anticoagulant (4%)

Table 2.

Percent of incomplete or potential lupus patients treated with various immunomodulatory medications.

	Brigham and Women’s Hospital N=161*	Spanish Rheumatology Society Lupus Registry N=345*
Hydroxychloroquine, %	65	ns+
Other antimalarial, %	1	69.4‡
Corticosteroids, %	35	69.1
Oral corticosteroids, %	35	ns
IV corticosteroids, %	1	ns
Azathioprine, %	2	13.2
Mycophenolate mofetil, %	2	3.7
Cyclophosphamide, %	1	6.3
Methotrexate, %	9	13.8
Sulfasalazine, %	6	0
Rituximab, %	1	1.6
Other biologics, %	3	ns
Leflunomide, %	ns	2.5
IV Ig, %	ns	3.1
Acetylsalicylic acid, %	ns	32.6
Oral anti-coagulants, %	ns	11.6
Statins, %	ns	18.8
Angiotensin-converting, enzyme inhibitors, %	ns	15.4
Anti-osteoporotic agents, %	ns	17.5
Plasmapheresis, %	ns	0.3
Dialysis, %	ns	0.6
Diuretics, %	ns	11.5

^*Please see text and references for additional details about the Brigham and Women’s Hospital (15) and Spanish Rheumatology Society Lupus Registry (13) cohorts.

⁺ns, not specified.

^‡Hydroxychloroquine was reported under the category of antimalarials.

Recent retrospective studies also found organ damage in ILE. A chart review of 15 ILE patients in the Dallas Regional Autoimmune Disease Registry showed significantly lower Systemic Lupus International Collaborating Clinics (SLICC)/ACR Damage Index (SDI; range of 0–2) scores in ILE patients (0.67±0.32) compared to SLE patients (1.67±0.17), and no renal damage was observed in ILE patients. However, other forms of organ damage did occur, as 5 ILE patients had SDI scores >0 (16). A total of 77 ILE patients, who had one to three 1997 ACR-SLE criteria, were recruited from inpatients at the Peking Union Medical College Hospital (14). The most prevalent criteria were ANA (97.4%), hematologic involvement (51.9%), immunologic disorder (29.9%), arthritis (20.8%), and serositis (20.8%) (14). The average SLE disease activity index (SLEDAI) score was 6.61±0.51, reflecting the presence of some increased disease activity. However, 18 ILE patients (23.4%) had SLEDAI scores of 10–14, and 4 patients (5.2%) had SLEDAI scores >14, indicating severe disease activity. The mean SDI score was 0.66±0.8, with a score of 0 in 36 patients (46.8%), 1 in 31 patients (40.3%), and 2 in 10 patients (13.0%). The most common form of organ damage was pulmonary arterial hypertension (22.1%), followed by renal (11.7%), neurological (10.4%), and peripheral vascular (9.1%) damage (14).

Estimates are that 10–50% of ILE patients will progress to SLE and that most transitions occur within 5 years of ILE onset, suggesting that many of the patients in these studies will not go on to develop SLE (13, 14). However, the severity of their clinical manifestations and the accrual of organ damage suggest that incomplete lupus merits attentive follow-up and treatment, even in the absence of formal SLE classification (13).

Predictors of future SLE classification

Ideally, understanding early pathogenesis in the spectrum of patients who subsequently will or will not develop SLE would help identify individuals at the highest and lowest risk of transitioning to classified SLE. After a period of monitoring, patients at the lowest risk of transitioning may be considered for less frequent follow-up, with reduced testing and avoidance of potentially toxic medications. Conversely, individuals at the highest risk of transitioning may benefit from early intervention, particularly when treatment is started before formal classification. In a retrospective study of 130 US military personnel who later met ACR-SLE classification criteria, pre-diagnosis treatment with hydroxychloroquine delayed the onset of classifiable SLE and reduced the number of autoantibody specificities at and after diagnosis (17). Therefore, there is great interest in developing a therapeutic strategy that can be implemented before SLE classification and identifying patients who would most benefit from early intervention.

Studies leveraging the large, longitudinal Department of Defense Serum Repository have shown that for many patients, SLE classification is preceded by a period of autoantibody positivity and other immune dysregulation, even in the absence of clinical symptoms (18, 19). Indeed, in a retrospective study, 104 of 130 US military personnel (80%) met at least one ACR criterion prior to SLE diagnosis (18), suggesting that a large percentage of patients may be identifiable before meeting full ACR-SLE classification criteria. Prospective, longitudinal studies of autoantibody-positive healthy individuals could potentially reveal some of the earliest biomarkers that predict future transition to SLE, but this approach is obstructed by the low percentage of autoantibody-positive healthy individuals who go on to develop SLE (20–23). Other studies have followed individuals with ILE, UCTD, or similar diagnoses to assess clinical characteristics, demographic information, or biomarkers that might distinguish the patients who will transition to SLE. Among individuals with UCTD, those with ANA-homogeneous pattern, anti-dsDNA, anti-Sm and aCL are at higher risk for transitioning to SLE, as are those with multiple autoreactivities or multiple clinical features of SLE (24–26).

A prospective, longitudinal study of 51 European women with UCTD showed that differences in T cell subsets are also associated with the transition to systemic autoimmune diseases (27). Over a mean follow-up of 2.05 years, 5 patients progressed to SLE, 7 to rheumatoid arthritis, 4 to mixed connective tissue disease, and 3 to systemic sclerosis. At baseline and follow-up, the number and frequency of Th17 cells were increased in UCTD patients compared to healthy controls, and in patients who developed systemic autoimmune diseases compared to patients who did not. In addition, the ratios of Th17 cells to natural regulatory T cells and Th17 cells to type 1 regulatory T cells were increased in UCTD patients compared to healthy controls, and in patients who developed systemic autoimmune diseases compared to patients who did not (27). Consistent with the emerging paradigm that an imbalance of inflammatory and regulatory factors contributes to the onset of autoimmune disease, these results suggest that altered T cell frequencies could help identify patients with impending autoimmune disease, although this may not be specific to SLE. In another study, 94 patients with UCTD were recruited in Cali, Colombia. In one year of follow-up, 13.8% of patients transitioned to a specific CTD: 8.5% to SLE, 4.2% to Sjogren’s syndrome, and 1.1% to rheumatoid arthritis. Patients who transitioned to a specific CTD were more likely to have arthritis, Raynaud’s phenomenon, and photosensitivity at the initial visit (28).

In a recently published study, all patients seen for potential SLE at the Brigham and Women’s Hospital were prospectively enrolled in the Brigham and Women’s Hospital Lupus Registry and characterized through retrospective chart review (15). Potential SLE was identified based on the presence of one to three ACR-SLE criteria and the opinion of a board-certified rheumatologist. At the initial consultation, 88.3% of patients had positive ANA, 53% had arthritis, 17% had anti-dsDNA, 14% had malar rash, and 2% had proteinuria or urinary casts (15). After a mean of 6.3 years of follow-up, 21% of patients had definite SLE, 18% did not have SLE, and 61% remained potential SLE patients. Patients who had definite SLE were significantly younger at follow-up (46.5 years) than patients without SLE (49.5 years) or with potential SLE (53.2 years). Consistent with other studies, oral ulcers (OR 2.40, 95% CI 1.03–5.58), renal manifestations (16.20, 1.63–161.02) and anti-dsDNA (2.59, 1.25–5.35) at the initial consultation independently predicted the development of SLE (15).

Two recent studies specifically addressed the possibility that autoantibodies could distinguish patients who will transition to SLE. A study of 60 Korean ILE patients with two or three ACR-SLE criteria considered anti-CRP as a potential predictor of transition. During the mean follow-up of 3.3 years, 15% of ILE patients transitioned to SLE, but anti-CRP levels were no different between patients who transitioned and those who did not. Other potential predictors of transition were not reported (29). Another group used an unbiased library of synthetic peptoids to determine whether differences in antibody reactivity might differentiate SLE patients, ILE patients, and ANA negative or ANA positive healthy individuals with no family history of autoimmune disease (30). Subjects were chosen from the Dallas Regional Autoimmune Disease Registry, and ILE was defined as the presence of an antinuclear antibody and one or two other criteria. Compared to pooled sera from healthy controls, pooled sera from SLE patients or ILE patients showed substantially greater reactivity to three peptoids, and analysis of individual sera showed that peptoid binding was due to distinct sets of autoreactive antibodies in SLE patients, ILE patients, and healthy controls. Therefore, this approach may prove useful in identifying patients who will transition to SLE. However, longitudinal studies are required to assess how serum reactivity to these peptoids changes during the progression from ILE to SLE (30).

Potential avenues for early intervention

As new research enables the reliable identification of individuals who are most likely to transition to SLE, the pre-classification period may become an ideal window for impeding the progression of autoimmunity and immune dysregulation that leads to clinical SLE (31). The immune pathways that are dysregulated during early pathogenesis, such as T helper/T regulatory pathways (27), may be amenable to therapeutic intervention. In addition, peroxisome proliferator-activated receptor γ (PPAR-γ) agonists, which are used clinically to enhance insulin sensitivity, have been found to reduce disease progression in mouse models of SLE when treatment is started before disease onset (32–34). In MRL/MpJ-Faslpr/2J and gld.apoE^−/− lupus-prone mice, the PPAR-γ agonist rosiglitazone reduces lymphadenopathy, serum ANA titer, anti-dsDNA antibody levels, and renal disease when treatment is started before the onset of disease (33). However, rosiglitazone treatment initiated after disease onset has no effect on these clinical manifestations in MRL/MpJ-Faslpr/2J and gld.apoE^−/− and may exacerbate kidney disease in MRL/MpJ-Faslpr/J mice (33). Finally, Lupus-prone B6.Sle1.Sle2.Sle3 mice given metformin with the glucose metabolism inhibitor 2-deoxy-d-glucose during the early stages of disease showed significant reductions in splenomegaly, production of anti-dsDNA IgG and ANA, and immune complex deposition in the kidneys. These changes correlated with downregulated metabolism in the effector T cells of treated vs. untreated lupus-prone mice (35). These studies suggest that, with additional study, medications currently approved for regulating metabolism may be helpful for early lupus intervention.

Plasmacytoid dendritic cells (pDCs) are another potential target for early intervention (36). Male BXSBxBDCA2-DTR transgenic mice were transiently depleted of pDCs from 8 to 11 weeks of age. At 11 weeks of age, treated mice showed impaired T cell activation, altered B cell populations, lower abundance of serum IgG and IgM, and reduced antibody reactivity to several autoantigens, including dsDNA, ssDNA, dsRNA, U1-snRNPs, La/SSB, chromatin, and histones (36). In addition, the IFN α/β gene signature, which is associated with SLE in humans, was attenuated in these mice. Interestingly, these effects were still observed at 19 weeks of age (8 weeks post-depletion), even though pDC populations were equal in untreated and previously treated mice. In addition, mice with previous pDC depletion showed reduced glomerulonephritis, demonstrating that early, transient pDC depletion had a sustained impact on disease pathogenesis (36). Together, these results suggest that delineating the unique pathogenetic mechanisms during the pre-classification period is an essential and promising step toward developing effective early intervention strategies for SLE.

Terminology for the pre-classification periods of SLE

The pre-classification period of SLE includes individuals with increased genetic risk of developing SLE but no ACR criteria, as well as individuals with multiple autoantibodies and clinical features suggestive of SLE but insufficient for SLE classification. The entire pre-classification period in individuals who will subsequently develop classified lupus has been called “pre-SLE.” In a strict sense, the related term “preclinical lupus” describes a period of immune dysregulation before the onset of clinical manifestations. Importantly, these terms can only be applied retrospectively after SLE diagnosis, since many individuals with features of SLE do not go on to develop lupus. Additional terms for use prior to diagnosis include “latent lupus,” typically indicating the presence of one or two ACR criteria plus selected minor criteria (4), and “incomplete lupus erythematosus” (ILE), indicating the presence of fewer than four ACR criteria (5). Other terms, including “incipient lupus” (4), “lupus-like” or “probable lupus” (37), and “overlap syndromes” (38), have been used historically but have fallen out of favor. “Undifferentiated connective tissue disease” (UCTD) is used more broadly to describe disease manifestations suggestive of connective tissue disease, but not specifically diagnostic for any given systemic autoimmune rheumatic disease (39). The first one to three years after symptom onset may be distinguished as “early UCTD” because transitions from UCTD to SLE or another defined CTD usually occur soon after the UCTD diagnosis (12). Several reviews focus on various aspects of the pre-classification period in SLE (40–42).

Changing concepts of the pre-classification period

Two ongoing conversations have the potential to reshape the way we think and talk about the period preceding SLE classification. First, there is a growing effort to develop methodical terminology for consistently describing various stages of SLE risk and development (12, 43, 44). Like the rheumatoid arthritis terminology developed by the European League Against Rheumatism (EULAR), a framework for SLE nomenclature could denote individuals with increased genetic and environmental risk (high-risk for SLE), preclinical autoimmunity and immune dysregulation (preclinical SLE), autoimmunity and select lupus clinical manifestations which are insufficient for classification (ILE), and classified SLE (43, 44). “Potential SLE” or “pre-classified SLE” have been proposed, but not widely adopted, as alternatives to “ILE” in individuals who do not meet SLE classification but have symptoms and autoimmunity convincing for lupus diagnosis (15). It is hoped that systematic terminology could help support a standardized clinical and experimental approach to pre-SLE, similar to that used for type 1 diabetes (45). Second, SLICC recently proposed new guidelines for SLE classification in hopes of increasing the numbers of patients participating in clinical trials who have convincing clinical evidence of SLE. The ACR guidelines, which were published in 1971 and revised in 1982 and 1997, classify SLE based on the presence of four criteria. The SLICC guidelines, which include several new or revised criteria, classify SLE based on the presence of four criteria (with at least one clinical and one immunologic criterion) or biopsy-confirmed lupus nephritis in the presence of antinuclear antibodies or anti-dsDNA antibodies (3). There is debate about whether clinical research would best be served by consistent use of ACR guidelines (46) or by adoption of the expanded criteria in the SLICC guidelines (47, 48). There is notable overlap between ACR and SLICC criteria, and studies in multiple ethnicities suggest that a large majority of patients meet ACR and SLICC classification criteria simultaneously. However, among 20 mixed connective tissue disease patients, 14 met ACR criteria, while 19 met SLICC criteria (49). Further, in those who do not meet ACR and SLICC criteria simultaneously, the SLICC criteria seem to allow for earlier classification (50, 51). Therefore, the definition of pre-SLE may change as nomenclature and SLE classification criteria evolve.

Conclusion

With recent advances in our understanding of potential lupus and the transition to classified SLE, the field is now poised to develop prospective natural history studies to assess clinical, immunological and serological biomarkers of disease onset and to elucidate mechanisms of disease that can be modulated by directed therapeutics. The development and validation of prognostic algorithms, reliable biomarkers, and directed therapeutics will allow for closer monitoring of patients at high risk of SLE transition and early intervention to delay or prevent full disease onset. Therefore, continuing research into the earliest stages of SLE may reduce end-organ damage and lupus-associated morbidity and mortality.

Key Points.

• Terminology for the pre-classification phases of SLE is evolving to reflect new concepts of disease pathogenesis and progression

• Patients with incomplete lupus could benefit from follow-up and treatment, even if they will not transition to classifiable SLE

• Therapeutic intervention during the pre-classification period could delay disease onset and reduce organ damage in patients at high risk of transitioning to classifiable SLE

• Additional studies are needed to confirm and validate whether certain autoantibodies, immunological changes, and clinical manifestations can distinguish patients who will transition to classified SLE

Abbreviations

ACR

American College of Rheumatology

EULAR

European League Against Rheumatism

ANA

antinuclear antibodies

ILE

incomplete lupus erythematosus

PPAR-γ

peroxisome proliferator-activated receptor γ

pDC

plasmacytoid dendritic cell

SLE

systemic lupus erythematosus

SLEDAI

SLE disease activity index

SLICC

Systemic Lupus International Collaborating Clinics

UCTD

undifferentiated connective tissue disease