Immune responses against endogenous nuclear antigens are characteristic of systemic lupus erythematosus (SLE), a highly pleiomorphic disease predominantly affecting young women of reproductive age. Genome-wide association studies have confirmed the importance of genes associated with the immune response as well as genes involved in endothelial function and tissue response to injury. Immune complexes, autoantibodies, complement, cytokines, endothelial injury and a thrombophilic state associated with antiphospholipid antibodies are important for mediating tissue dysfunction. If not treated promptly, a significant proportion of patients—especially those with more aggressive disease—accumulate irreversible damage. During the past decade, novel combinations of immunosuppressive drugs and biologicals have been added to the therapeutic armamentarium. At the same time, the emphasis in the management of lupus has shifted from individual drugs to a strategy that aims at early, sustained remission tailored to disease manifestations and severity with the lowest possible toxicity. Infections and accelerated atherosclerosis (attributed to both traditional and non-traditional risk factors) and thrombosis-related clinical events (including arterial, venous and pregnancy loss) represent a major challenge in the management of the disease. To avoid fragmentation and optimise medical care, evidence and expert-based recommendations have been developed. For the future the authors predict a new taxonomy on the basis of mechanisms rather than clinical empiricism, leading to targeted therapy.
Statistics from Altmetric.com
The period 2000–10 represents a ‘golden decade’ for rheumatology thanks to major advances in understanding and the introduction of biological therapies. During this period, there were many discoveries related to systemic lupus erythematosus (SLE), the prototypic systemic autoimmune disease. Modern genetics confirmed the involvement of essentially all components of the immune system in disease pathogenesis and identified additional factors related to tissue targeting and vulnerability to immune attack and injury. The nucleic acid targets of antinuclear antibodies (ANA)—a distinct feature of SLE—were linked to impaired clearing of apoptotic cell products and shown to trigger immunological activation and initiate novel cellular immune pathways. The orderly appearance of distinct autoantibodies up to 9 years before disease onset,1 indicative of a preclinical phase of disease, has created possibilities for prediction and prevention of disease. At the clinical level, the studies of autoantibodies, including antiphospholipid antibodies (APA), have defined subtypes of the disease with unique phenotypes, mechanisms, prognosis and requirements for management. Increased attention to comorbidities associated with SLE, particularly atherosclerosis, has increased importance of primary and secondary prevention. New immunosuppressive drugs and success of the first biological agent—after several failed trials—with potential therapeutic benefit have created optimism in the community. Immunosuppressive therapy in lupus can put the disease in remission in over two-thirds of patients with nephritis,2 3 a target that has yet to be reached in other autoimmune rheumatic diseases. In this review we discuss some of these developments and their implications for the care of SLE patients. This is not a comprehensive description of advances in the disease; rather, we focus on evolving concepts and major shifts in understanding pathogenesis and management paradigms.
Pitfalls in diagnosis and ANA testing
Lupus remains a clinical syndrome with validated features. The American College of Rheumatology (ACR) classification criteria, developed in the context of epidemiological studies to ensure that SLE patients do in fact have a similar disease, are also commonly used as diagnostic in the clinic. There are several caveats, however, in their general use, and strict adherence may miss patients' lupus-like disease. ACR criteria for SLE may exclude patients with early or limited disease and limit their access to treatments. These criteria are not weighted for specificity, sensitivity or severity. Data from tertiary centres suggest that only two-thirds of patients referred for lupus fulfil the ACR criteria, approximately 10% of these have features of lupus but do not fulfil criteria, and 25% have fibromyalgia-like symptoms and positive ANA test but never develop lupus.4
Undifferentiated connective tissue syndromes account for 10–20% of referred patients, 10–15% fulfilling classification for SLE 5 years later. Factors that predict evolution to SLE are young age, alopecia, serositis, discoid lupus, positive Coombs and anti-Sm or anti-DNA antibodies.5 Incomplete lupus describes patients who present with a constellation of symptoms suggestive of SLE, but do not qualify by clinical ‘intuition’ or classification criteria as having classic SLE.6 These patients usually present with one to two ACR criteria and other features not included in the classification criteria. Most of these patients do not develop SLE or in case they do, it is usually mild and rarely involves major organs. We await biomarkers to improve diagnosis and prognosis.
Although it is often said that at least 95% of SLE patients are ANA positive by immunofluorescence, the sensitivity of ANA testing may be as low as 70%.7 This is particularly true in laboratories that employ enzyme immunoassays or other automated assays that display marked inter-manufacturer variation in performance and the reported sensitivity against positive immunofluorescence-ANA with titre at 1:160 ranges from 70% to 98%.8 9
Prognostic markers and the role of autoantibodies
Analysis of large SLE cohorts has defined clusters of autoantibodies associated with distinct features of disease severity.10 11 Serum anti-dsDNA titres—measured by the Farr assay but not by ELISA—have been correlated with nephritis (OR typically in the range of 1.6–1.8), progression to end-stage renal disease (OR 4.1) and increased disease severity, damage or poor survival.12 13 In the LUMINA cohort, anti-dsDNA antibodies at baseline were associated with shorter time to initial damage (Systemic Lupus International Collaborating Clinics (SLICC)/ACR) damage index (SDI) >0) (HR 1.8). APA are strongly associated with features of the antiphospholipid syndrome (APS) (arterial/venous thrombosis, fetal loss, thrombocytopenia) (OR ranging 2.8–5.4), central nervous system involvement (OR ranging 3.1–4.5 for any neuropsychiatric involvement/damage, 3.3–22.2 for cerebrovascular disease), severe renal involvement (RR 2.2 for progression to end-stage renal disease), damage accrual (OR ranging 1.9–2.8) and death.13 14 Anti-C1q antibodies correlate with renal disease,15 and antibodies to extractable nuclear antigens (anti-Ro/La/Sm/RNP) have been associated with mucocutaneous involvement and less severe nephropathy in most studies.11 It has been suggested that these clinicoserological subsets correspond to distinct genetic variants of susceptibility/severity genes, but this requires further investigation.16 Importantly, autoantibodies have not been found to be informative across SLE patients in all clinical settings and no single or combination of factors has emerged to predict outcomes accurately.
Clinical course: disease activity and damage
Clinical activity of SLE may wax and wane, but persistent, smouldering systemic inflammation leads to accrual of organ damage (figure 1). Validated global and organ-specific activity indices are used in the clinical evaluation of SLE.17 These indices, developed in the context of long-term observational studies, may predict damage and mortality, and reflect longitudinal changes in disease activity. However, there is little evidence that any of these disease activity markers can be used as surrogates of treatment effect in trials. Also, the minimum clinically meaningful change in these markers to trigger alterations in therapy remains empirical at present. Novel biomarkers such as type I interferon (IFN)-inducible genes18 and circulating subpopulations of B lymphocytes,19 have been correlated with SLE severity but need further validation that can guide treatment.
Damage describes irreversible changes in organ integrity and function and its prevention is a goal for new therapy in SLE. Accumulation of damage is a function of disease activity over time, side-effects of treatment and/or comorbid conditions, and is clearly linked to morbidity and mortality. Despite improved survival rates (92–98% at first 10 years), SLE patients still have two to fivefold increased risk of death compared with the general population.20 Early damage and mortality (during the first 1–2 years after diagnosis) is mostly related to disease, whereas later damage, namely atherosclerosis, infections and malignancy, is usually associated with complications of long-standing disease and treatment with immunosuppressive agents.21 22 The SLICC/ACR damage index is a validated instrument specifically designated to ascertain damage in SLE (figure 1). Several independent studies have shown that early acquisition of damage is a poor prognostic factor.23 24
Comorbidities in SLE
SLE patients have increased cardiovascular morbidity unexplained by traditional risk factors.25 The presence of atherosclerosis, defined by carotid ultrasound as the presence of plaque and/or increased intimal-medial thickness is accelerated and more prevalent in SLE. Premature atherosclerosis is associated with longer lupus disease duration, higher damage score and less aggressive immunosuppressive therapy.26 27 There is no consensus on who to screen or which screening modality to use to define preclinical atherosclerosis. In the absence of controlled trials to test the efficacy of specific primary or secondary prevention strategies (eg, treatment with aspirin, statins, antimalarials) in SLE, clinicians should assess cardiovascular risk in individual patients, aggressively treat modifiable risk factors and control disease activity to minimise the atherogenic effects of inflammation.
Infections account for 20–40% of all deaths in lupus. Susceptibility to infections has been associated with disease duration and activity, leucopenia, use of glucocorticoids (moderate/high dose) and immunosuppressants and certain manifestations, particularly nephritis.28 Vaccination against influenza and pneumococcus is essential. Herpes zoster vaccination, currently recommended for individuals older than 60 years, may also be considered but data on efficacy and safety in younger SLE patients are not available. Despite concerns for precipitating flares, inactivated vaccines are safe but immune responses may be limited in patients on immunosuppressive treatments. Live vaccines (measles, mumps, rubella, polio, varicella, zoster vaccine, smallpox) are contraindicated in the immunosuppressed or patient on 20 mg or more prednisone daily.
Certain cancers occur more frequently in SLE, including haematological malignancies (particularly non-Hodgkin's lymphoma), cervical, breast and lung cancer.29 Women with SLE are at increased risk of human papillomavirus (HPV) infections (including high-risk and multiple HPV type infection) associated with cytology abnormalities, and this risk correlates with increased inflammatory burden (increased damage at baseline and use of cyclophosphamide).30 31 Therefore, using the combination of cervical cytology plus HPV DNA testing may be preferred to screen SLE women older than 30 years. Vigilance in cancer preventive strategies at least similar to those of the general population is currently recommended, including HPV vaccination.
Management of SLE: overriding principles and strategy
Lupus is a life-long disease of variable severity. Persistent activity or flares interspersed among periods of remission lead to accumulation in disease and treatment-related damage. Accordingly, therapeutic strategies with multiple targets should be aimed at reducing overall burden of systemic inflammation. Such approaches require the capacity for early diagnosis, precise definition of disease activity and flare, stratification according to severity of involvement, use of drugs to induce remission promptly and prevent flares, and prevention and management of comorbidities. Treatment of nephritis and other severe lupus manifestations commonly involve an intensive induction of remission phase and a less intensive maintenance phase, which may be accomplished with single agents alone, in combination or sequentially. Drugs that have demonstrated efficacy in lupus include hydroxychloroquine, methotrexate, azathioprine, mycophenolate mofetil (MMF) and cyclophosphamide.32 The role of thalidomide, leflunomide, calcineurin inhibitors (cyclosphorine and tacrolimus) and intravenous immunoglobulin in SLE treatment is not yet clear. Rituximab (anti-CD20 monoclonal antibody) has been effective in treating severe, refractory lupus manifestations (especially autoimmune cytopaenias) in case series,33 but failed to control disease in randomised controlled trials (RCT).34 35 The emerging use of biologicals holds great promise with the first successful report with belimumab (fully human monoclonal antibody that binds to and inhibits human B lymphocyte stimulator).36 To ensure compliance, treatment has to consider profile and preferences of the patients. Because the care of lupus patients may involve several medical specialties and in order to avoid fragmentation in their care, the European League Against Rheumatism (EULAR) Task Force on SLE has developed management recommendations based upon evidence and expert opinion.37
Nephritis remains the single most important cause of morbidity and a major cause of mortality in lupus. The revised renal classification criteria by the International Society of Nephrology and the Renal Pathology Society in 2004 draws attention to quantitative and qualitative morphological data of the glomeruli, tubulo-interstitial compartment and vasculature, and distinguishes global from segmental disease.38 Stratification based on renal pathology and patient demographic, clinical and laboratory characteristics, enables identification of patients at high risk of renal dysfunction, who may benefit from aggressive therapy (figure 2).39
The efficacy of MMF in proliferative lupus nephritis both as an induction and as a maintenance regimen has been tested in RCT.40 In the largest induction study response rates were 56% for MMF versus 53% in the cyclophosphamide group.41 Unfortunately to this point, follow-up of these studies ranges from short (≤1 year) to medium (≤5 years) term with longer follow-up (≥10 years) not available at present as is the case with studies with high or low-dose cyclophosphamide.42 43 Notwithstanding these limitations, an increasing number of physicians currently use MMF as first therapy for most cases of proliferative lupus nephritis because of its favourable toxicity profile while reserving cyclophosphamide for more severe cases. Whether MMF is superior to azathioprine as a maintenance regimen is not clear, with two recent studies (Mycophenolate Mofetil Versus Azathioprine for Maintenance Therapy of Lupus Nephritis (MAINTAIN),44 Aspreva Lupus Management Study (ALMS))45 reaching different conclusions. This may be due to differences in racial contribution (MAINTAIN involved only Europeans whereas ALMS a mixture of patients some of which are non-Caucasians who may be more likely to respond to MMF) and the larger size of the ALMS study which showed superiority. Of interest, a head-to-head comparison of azathioprine with pulse intravenous cyclophosphamide in European patients demonstrated good efficacy for azathioprine, albeit not comparable with cyclophosphamide.46 More recent studies thus support a role of azathioprine for milder cases of lupus nephritis or as a long-term maintenance regimen. Of note, studies on azathioprine and MMF involved a mixed population of patients ranging from mild to severe diseases. Studies involving exclusively patients with severe lupus nephritis exist only for intravenous cyclophosphamide.47
One-third to one-fourth of patients with moderate to severe proliferative lupus nephritis will relapse after partial or complete remission arguing for the need to identify safe long-term treatments or better induction therapy.48 Rituximab may offer some benefit in these cases or in patients who cannot be treated with immunosuppressive agents based on evidence from series of cases.49 Patients with membranous disease may benefit from glucocorticoids alone or in combination with cyclosporine or cyclophosphamide.50
Neuropsychiatric manifestations in SLE involve the central and the peripheral nervous system and range from overt manifestations such as stroke, seizures and psychosis, to more subtle abnormalities of cognitive function. Less than 40% of neuropsychiatric events can be attributed to lupus, whereas the remaining represent complications of the disease or its therapy, or may be due to infections, metabolic abnormalities and drug adverse effects.51 Multiple pathological mechanisms are implicated in neuropsychiatric systemic lupus erythematosus (NPSLE), including antiphospholipid or other autoantibody-mediated vascular or neuronal injury, intrathecal production of inflammatory mediators and accelerated atherosclerosis. Despite advances in the understanding of the immunopathogenic and clinical aspects of SLE, NPSLE remains a diagnostic and therapeutic challenge. Difficulties include the correct attribution of neuropsychiatric syndromes to SLE, the selection of proper diagnostic examinations, and the optimal management of NPSLE in view of limited clinical trial data. To facilitate the management of these patients, EULAR has recently developed recommendations using an evidence-based approach followed by expert consensus (table 1).52
APA and APS
APA (anticardiolipin, anti-β2-glycoprotein 1 and lupus anticoagulant; LAC) are found in 30–40% of SLE and are associated with increased risk of thrombo-occlusive incidents. It has been suggested that thrombotic risk assessment may improve by combining APA with selected prothrombotic markers such as P-selectin and factor VII levels.53 The combination of vascular thrombosis and/or obstetric morbidity (defined as unexplained deaths of morphologically normal fetuses at 10 or more weeks of gestation, newborn premature losses of morphologically normal fetuses to 34 weeks of gestation or before due to severe preeclampsia or eclampsia or placental insufficiency, three or more consecutive unexplained spontaneous abortions before 10 weeks of gestation) and persistently positive APA (>40 GPL (IgG anticardiolipin units) or MPL (IgM anticardiolipin units) and/or positive LAC) measured at least 12 weeks apart defines APS.54 55 Several clinical features are more common in SLE patients with APS such as thrombocytopenia, livedo reticularis, heart valve disease, pulmonary hypertension, central nervous system and renal disease, but these manifestations have little specificity.10 A distinct type of small-vessel vaso-occlusive nephropathy with histological features of thrombotic microangiopathy and chronic vascular lesions, termed APS nephropathy, would be suspected in APA-positive patients who present with hypertension (often severe), proteinuria (usually mild to moderate), haematuria and renal impairment.56 These patients may benefit from aggressive immunosuppressive therapy along with anticoagulation. Based upon small controlled trials57 and in-vitro data suggesting anti-proliferative and other suppressive effects of MMF on vascular smooth muscle and endothelial cells, some authors have claimed that MMF may be superior to cyclophosphamide in such patients, but these data need to be validated.
Thrombosis as a consequence of APS requires anticoagulation therapies to prevent recurrent events but the intensity of such therapies remains controversial in view of the paucity of properly designed RCT. Patients with definite APS and first venous or arterial non-cerebral/non-coronary thrombosis should be treated with oral anticoagulation at a target international normalised ratio of 2.0–3.0,58 although some experts recommend higher-intensity anticoagulation following arterial thrombotic events. Data from APA and stroke studies and large RCT in the general population suggest that antithrombotic therapy is not superior to antiplatelet therapy for secondary thromboprophylaxis after non-cardioembolic stroke or transient ischaemic attack.58 Because many patients studied were older people and had low titres of APA, determined at one time point, the conclusions may be limited to these populations. Acute coronary artery syndromes and myocardial infarction should be treated according to the evidence base for the general population. For patients with recurrent thromboses, anticoagulation should target international normalised ratio of 3.0–4.0, especially if they have high-risk profile of APA (LAC or anticardiolipin IgG at higher titres, or anti-β2-GP1 plus LA or anticardiolipin).59 Additional atherothrombotic risk factors should be aggressively controlled. A small proportion (<1%) of patients may have a potentially life-threatening variant of APS, catastrophic APS, which is characterised by multiple small-vessel thromboses resulting in multi-organ failure.60
Pregnant SLE–APS patients are at increased risk of complications including maternal thrombosis, recurrent spontaneous abortions before 10 weeks gestation and late adverse pregnancy outcomes such as fetal death, pre-eclampsia, fetal growth restriction and preterm birth.61 For women with APS and a history of pregnancy complications and or thrombosis, combined therapy of heparin and aspirin during pregnancy and postpartum has been reported to reduce pregnancy loss by up to 54%,62 but further studies are needed to determine appropriate and safe treatment for these women. Specifically, comparisons of unfractionated heparin to low molecular weight heparin are lacking and the contribution of aspirin is unclear. A beneficial effect of low-dose aspirin in primary prevention of thrombotic events and spontaneous abortion in SLE with persistently positive moderate-to-high titres of APA has been suggested by some,63 64 but not all,65 studies. Of note, neither aspirin combined with low molecular weight heparin nor aspirin alone improved the live birth rate, compared with placebo, among women with unexplained recurrent spontaneous abortion.66 The emerging role for APA in activation of complement, induction of tissue factor expression and impairment of placental development may lead to new approaches to prevent and treat APS.67–69
Clinical trials in SLE
Hydroxychloroquine, corticosteroids and aspirin are the only US Food and Drug Administration approved drugs in lupus; no new drug has been added in the last 30 years. Off-label use of azathioprine, MMF, cyclophosphamide, methotrexate, leflunomide and cyclosporine is common practice either as steroid-sparing agents or for severe lupus. Studies of novel targeted drugs including LJP-394 toleragen and dehydroepiandrosterone were unsuccessful. RCT with MMF failed to support claims of superiority to cyclophosphamide in earlier studies. Biological therapies of proven efficacy in rheumatoid arthritis such as abatacept (CTLA4Ig) and rituximab failed to demonstrate efficacy in lupus RCT. Fortunately, a new biological agent, belimumab, overcame the perceived ‘curse’ in lupus trials and showed efficacy by a composite responder index, albeit with a small treatment effect.36
We do not know why recent trials of biologicals failed to meet their primary endpoints. Potential explanations include effects of background therapies, criteria for selection of patients and measures of outcomes. Notwithstanding the inherent difficulties in trial design, we remain optimistic that new treatments and therapies with demonstrated efficacy in other diseases will reach lupus. To facilitate SLE drug development efforts, the EULAR Task Force on SLE has used a combination of research-based evidence and expert consensus to create points to consider in the design of SLE trials.17 Table 2 summarises targeted biological or small-molecule therapies in SLE based upon a better understanding of disease pathogenesis.
What causes SLE? Aetiology, pathogenesis and clinical implications
Genetic and environmental factors
The aetiology of SLE has genetic and environmental components with female sex strongly influencing pathogenesis. These factors lead to an irreversible break in immunological tolerance manifested by immune responses against endogenous nuclear antigens. Genome-wide association studies have confirmed the importance of genes associated with immune response and inflammation (HLA-DR, PTPN22, STAT4, IRF5, BLK, OX40L, FCGR2A, BANK1, SPP1, IRAK1, TNFAIP3, C2, C4, CIq, PXK), DNEA repairs (TREX1), adherence of inflammatory cells to the endothelium (ITGAM) and tissue response to injury (KLK1, KLK3).70 The risk may be influenced by epigenetic effects such as DNA methylation and post-translational modifications of histones, which can be either inherited or environmentally modified.71
How does lupus start?
Lupus starts with a preclinical phase characterised by autoantibodies common to other systemic autoimmune diseases and proceeds with a more disease-specific clinically overt autoimmune phase. The role of major stress and environmental factors, namely drugs, sun exposure and infections—especially Epstein–Barr virus infection72—precipitating factors in a subset of patients is well documented. Figure 3 summarises key pathogenetic events in the disease. Increased amounts of apoptosis-related endogenous nucleic acids stimulate the production of IFNα and promote autoimmunity by breaking self-tolerance through activation of antigen-presenting cells. Once initiated, immune reactants such as immune complexes amplify and sustain the inflammatory response. In addition to the autoimmune and inflammatory components, endothelial injury and a thrombophilic state associated with APA are important for mediating tissue dysfunction and injury in a variety of organs.
Disease mechanisms and tissue damage
Immune complexes and complement-activation pathways are essential for effector function and tissue injury. Tissue damage is mediated by recruitment of inflammatory cells, reactive oxygen intermediates, production of pro-inflammatory cytokines and modulation of the coagulation cascade. Autoantibody-mediated tissue injury has been implicated in NPSLE, in which antibodies reacting with both DNA and glutamate receptors on neuronal cells can mediate excitotoxic neuronal cell death or dysfunction.73 Locally produced cytokines such as IFNα and tumour necrosis factor contribute to affected tissue injury and inflammation. These mediators together with the cells producing them (macrophages, leucocytes, dendrite cells and lymphocytes) are the subject of investigation as potential therapeutic targets in lupus. Recent studies have also highlighted the role of locally expressed factors for the protection of tissues under immune attack. For example, defects in kallikreins may jeopardise the ability of lupus kidneys to protect themselves from injury,74 PD-1-ligand downregulates the activity of the infiltrating lymphocytes,75 and impaired regulation of complement amplifies vascular injury.76
Vascular damage in SLE has received increased attention in view of its relationship with accelerated atherosclerosis. Homocysteine and proinflammatory cytokines impair endothelial function and decrease the availability of endothelial precursor cells to repair endothelial injury. Production of pro-inflammatory high-density lipoproteins and a dysfunction of high-density lipoproteins mediated by antibodies have also been implicated in defective repair of endothelium.77 78 Moreover, pathogenic variants of ITAM alter its binding to ICAM-1 and may increase the adherence of leucocytes to activated endothelial cells. Impaired DNA degradation as a result of mutations of the 3′ repair exonuclease 1 (TREX1) and increased accumulation of single-stranded DNA derived from endogenous retro-elements in endothelial cells, may activate the IFN-stimulatory DNA response and direct immune-mediated injury to the vasculature.
Well-known inducers of lupus flares, such as stress, ultraviolet irradiation and infections—especially viral infections—through the activation of various sensors of the innate immune response could contribute to initiation or flaring of the disease. Epstein–Barr virus, a major risk factor for lupus, promotes IFNα production by plasmacytoid dendritic cells suggesting that elevated IFNα levels in lupus may be—at least in part—due to aberrantly controlled chronic viral infection. Acquired defects in Toll-like receptors (TLR) signalling has been reported to correlate with remission of SLE, while various oligonucleotide TLR antagonists or inhibitors of downstream signalling pathways, with the most notable being the antimalarials, which inhibit TLR signalling, are in use or under development in lupus. Antibodies to IFNα are in clinical trials with early indications of blocking the IFN-inducible gene expression. Inhibition of complement activation by heparin may be contributing to its beneficial effects in patients with APS. Genes related to the activation of the immune system in lupus may be useful as biomarkers for early diagnosis, prognosis, monitoring of disease activity and prediction of response to treatment.
The future of SLE: final perspective
Lupus made a fleeting appearance as a systemic disease late in the 19th century and acquired complete clinicopathological ‘status’ in the first half of the 20th century. A defining event in lupus pathogenesis and diagnosis was the description of the LE cells in the middle of the 20th century, soon to be followed with the description of antinuclear and extractable nuclear antigens. The field has witnessed major advancements in defining risk factors and phenotypes, elucidating pathogenesis and treatment. These scientific advances have broad implications for autoimmune diseases and regulation of inflammation. As we approach the second decade of the 21st century, lupus has at last shown signs of yielding to biological therapies and revealing its heterogeneity and complexities.
The authors thank Dr Michael D Lockshin for critically reviewing the manuscript.
Funding This study was supported by the Hellenic Rheumatology Society.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.