Unmet medical needs in systemic lupus erythematosus

Systemic lupus erythematosus (SLE) is an autoimmune disease of diverse manifestations, with onset usually in young women in the third to fourth decade of life. The chronic nature of this relapsing remitting disease leads to organ damage accrual over time. Mortality and morbidity are increased in patients with SLE compared with the general population. Therapeutic advances over the last few decades have led to significant improvements in patient outcomes. Five-year survival has improved to over 90% from a low of 50% in the 1950s. However, multiple aspects of the management of SLE patients are still far from optimal. Early diagnosis remains a challenge; diagnostic delays leading to delay in definitive treatment are common. Monitoring treatment remains problematic due to the paucity of sensitive biomarkers. Current treatment regimens rely heavily on corticosteroids, even though corticosteroids are well known to cause organ damage. Treatment of refractory disease manifestations such as nephritis, recalcitrant cutaneous lesions and neurological involvement require new approaches with greater efficacy. Cognitive dysfunction is common in SLE patients, but early recognition and adequate treatment are yet to be established. Premature accelerated atherosclerosis remains a leading cause of morbidity and mortality. Fatigue is one of the most disabling symptoms, and contributes to the poor quality of life in patients with SLE. Ongoing research in SLE faces many challenges, including enrollment of homogeneous patient populations, use of reliable outcome measures and a standard control arm. The current review will highlight some of the outstanding unmet challenges in the management of this complex disease.

Rheumatology (ACR) classifi cation criteria, although not designed for diagnostic purposes, have been used by rheumatologists for almost three decades [1]. Th e Systemic Lupus International Collaborating Clinics (SLICC) group recently developed another set of revised criteria. Th ese new criteria include 17 variables derived by expert consensus (SLICC committee members) and statistical analysis, using real-life patient scenarios. Th e fi nal set of new criteria was then validated in another group of SLE patients and controls. Th e control group comprised other autoimmune diseases, which may have overlapping features with SLE.
SLICC criteria require that at least one clinical criterion and one immunologic criterion be present, with a total of four criteria, to have a classifi cation of SLE. Under this new classifi cation, lupus nephritis by biopsy (in the presence of SLE autoantibodies) is suffi cient for classi fication [2]. Using expert consensus as the gold standard, the revised criteria demonstrated greater sensitivity (97% vs. 83%, P <0.0001) but less specifi city (84% vs. 96%, P <0.0001) than the current ACR criteria in the validation group [2]. Th ese criteria are clinically more relevant and will probably identify more patients with clinically defi ned lupus than using the current ACR criteria. However, one should stress that these criteria are primarily meant for classifi cation of patient cohorts for research and their use for diagnostic purposes has to be carried out with caution. Some patients may initially present with insuffi cient features to fulfi ll the classifi cation criteria, termed 'incomplete lupus' or 'lupus-like disease' by some groups.

Biomarkers in systemic lupus erythematosus
SLE is the most diverse autoimmune disease, clinically and serologically. Genetic infl uences as well as epigenetic and environmental interactions probably play a role, perhaps dependent on the ethnic background of the individual. Classifi cation criteria may help but currently, there is no single parameter that is sensitive or specifi c enough to correctly identify or subtype all SLE patients. Similarly, the disease course can be variable with either intermittent fl ares or chronic activity. Levels of autoantibodies (anti-double-stranded (DNA) antibody) and com plement components represent serologic disease activity. Th ey are routinely used to monitor disease activity in most clinical settings, but their association with clinical activity has not been consistent in longi tu dinal studies [3].
Th ere is an unmet need for more sensitive and reliable biomarkers that can predict susceptibility, activity, severity and disease subtype in SLE. Multiple candidate markers have been proposed, including the type 1 interferon signature, B-lymphocyte stimulator and many others [4][5][6][7][8][9][10][11][12][13]. Th ese markers are currently only a research tool and not a single biomarker has been validated for clinical use to date. As SLE is a heterogeneous disease, a single biomarker is unlikely to be suffi cient. Rather, diff erent markers may provide information about specifi c disease aspects, as summarized in Table 1.

Management of refractory disease
Th ere is no clear defi nition of refractory disease in SLE, but it generally refers to patients who fail to respond to conventional treatments. In a heterogeneous disease such as SLE, the clinical situation may vary, depending on the disease manifestations and organ involvement. Although any clinical feature may become persistent and nonresponsive to therapy, the most concerning features are refractory lupus nephritis, scarring cutaneous disease and neuropsychiatric lupus (NPSLE).

Lupus nephritis
Renal involvement is a major cause of mortality and morbidity in SLE. A large proportion of patients, up to 60%, develop immune complex-mediated lupus nephritis during the course of their disease. Th e treatment of lupus nephritis has rapidly advanced over the last few decades. Glucocorticoids and cyclophosphamide were once considered the standard of care. Although eff ective in the majority of patients, cyclophosphamide was associated with serious adverse eff ects including infections, malignancy and infertility. Lower doses of cyclophosphamide and mycophenolate emerged as eff ective options for induction therapy, with better safety profi les [14,15]. Patients with diff erent ethnic backgrounds might have diff erential responses to cyclophosphamide versus mycophenolate [16]. Azathioprine and mycophenolate have been shown to be eff ective options for maintenance therapy in randomized trials [17,18]. Mycophenolate was superior to azathioprine in the ALMS trial, but not in a MAINTAIN trial [17,18].
Alternate approaches have been tried for patients failing to respond to these conventional treatments. Calcineurin inhibitors, cyclosporine and recently tacrolimus have shown promising results in patients unrespon sive to fi rst-line therapies, but need further evaluation in larger controlled trials [19][20][21]. B cells play a central role in the pathogenesis of lupus nephritis, making them a logical therapeutic target. Rituximab, a chimeric anti-CD20 antibody, effi ciently and reliably depletes CD20positive B cells. A large number of open-label studies documented the effi cacy of rituximab in refractory lupus nephritis [22]. Unfortunately, a large randomized controlled trial (LUNAR) did not show any signifi cant diff erences in outcomes with rituximab compared with placebo [23]. However, this trial excluded patients with refractory disease, the very subset in which evidence of benefi t was shown in open-label studies. Additionally, heavy background immunosuppression may have masked any benefi cial eff ect of rituximab [24]. Open-label data continue to be positive and a recent systematic review concluded that evidence for rituximab effi cacy in refractory lupus nephritis is strong, and another wellconducted trial may provide more answers [25,26].
Hydroxychloroquine deserves special mention in the treatment of SLE, including lupus nephritis. Hydroxychloroquine has been shown to improve response rates, decrease fl ares and improve survival [27,28]. Every SLE patient should receive hydroxychloroquine, unless intoler ant or contraindicated. Renal protective therapies such as angiotensin-converting enzyme inhibitors and angiotensin receptor blockers, strict control of blood pressure and serum lipids, are important adjuncts to therapeutic regimens for lupus nephritis [29,30].
Despite aggressive immunosuppressive and supportive therapies, induction of remission may be slow or partial, relapses remain common and progression to renal damage may occur. Th e incidence of end-stage renal disease attributable to lupus nephritis has not declined over the decades [31,32]. About 10 to 30% of SLE patients still progress to end-stage renal disease with associated morbidity and mortality. Multiple variables aff ect renal prognosis, including the disease severity, antibody associations, ethnicity, genetic background, socioeconomic status and concomitant co-morbidities [33,34].

Cutaneous lupus
Skin involvement occurs in 70 to 85% of SLE patients and includes acute, subacute and chronic cutaneous lesions. Discoid lupus is the most common form of chronic lesion, and may be the initial presentation of SLE in up to 10% of cases. Resistant discoid lesions may cause scarring of the aff ected skin areas and permanent scarring alopecia. Th is can lead to signifi cant disfi gurement, emotional distress, physical limitations and disability in a large number of patients. Although generally believed to be associated with less severe systemic disease, discoid lupus has also been associated with more damage accrual [35]. In addition, the risk of squamous cell carcinoma is increased in scarred areas. Th e conventional approaches include photoprotection, corticosteroids (topical, intralesional and systemic) and antimalarial agents [36]. Other topical agents such as calcineurin inhibitors and retinoids can benefi t some patients with refractory lesions. Multiple immuno suppres sants including methotrexate, azathioprine, mycopheno late, cyclophosphamide, dapsone, gold and thalidona mide have been tried for refractory disease [36]. Biologic agents including TNF inhibitors, efazulimab (anti-CD11 antibody) rituximab (anti-CD20 antibody) and tocilizumab (anti-IL-6) have been used in some refractory cases [37]. Data are limited to anecdotal reports, case series and open studies. A recent Cochrane review concluded that the evidence about therapies for discoid lupus, other than topical corticosteroids and antimalarials, was not conclusive [38]. Th e need for additional therapies for such a disfi guring manifestation of young people with SLE cannot be overemphasized.

Neuropsychiatric lupus
NPSLE remains one of the most challenging issues in the management of SLE patients. Aff ecting up to 30 to 40% of the patients, NPSLE includes diverse neurologic and psychiatric manifestations, from subtle cognitive defi cits to severe psychosis, seizures and strokes. Th e attribution of neuropsychiatric events to SLE is often challenging in the clinical setting. Th e ACR committee has described case defi nitions for 19 NPSLE syndromes, but the specifi city remains low [39,40]. Anti-ribosomal P antibodies have been associated with NPSLE, especially psychosis [41]. Measurement of cerebral spinal fl uid cytokine and chemokine levels remains a research tool [42]. Neuroimaging can be helpful but the sensitivity and specifi city remain quite low [43].
Th e management of NPSLE depends on the manifes tation and the likely predominant mechanism. Current strategies include the use of immunosuppressive therapies when the underlying mechanisms are pre dominantly infl ammatory. Anticoagulation and/or anti platelet therapy should be considered when anti phos pholipid antibodies are persistently positive in moderate to high titers. Non-SLE precipitation and aggra vat ing factors should be addressed. Cognitive dysfunction merits special mention. It is present in a majority of SLE patients, even newly diagnosed patients [44,45], and has also been associated with psychosocial factors such as depression, fatigue, anxiety and pain [45,46]. Although anti de pressants may benefi t some patients with depression and cognitive dysfunction, no treatment of proven benefi t exists for cognitive dysfunction in SLE [45,47].
NPSLE management remains problematic due to the lack of specifi c tools for diagnosis and attribution. Treatment options are limited to glucocorticoids and a few immunosuppressants (cyclophosphamide, azathioprine, mycophenolate, rituximab), with effi cacy mostly suggested by case reports and open-label studies. A European League Against Rheumatism task force recently published recom mendations on the management of NPSLE; it was felt that 'there is currently no good quality evidence to guide several diagnostic, primary prevention, therapeutic and monitoring decisions in NPSLE, emphasizing the need for further research' [48].

Atherosclerosis and systemic lupus erythematosus
SLE is associated with premature and accelerated atherosclerosis, contributing signifi cantly to the increased mortality and morbidity associated with the disease [49]. An increased frequency of conventional risk factors, such as hypertension, dyslipidemia and diabetes, has been noted in patients with SLE. Yet the excess risk cannot be fully explained by the traditional Framingham risk factors [50,51].
Multiple putative mechanisms have been proposed but the exact pathogenesis of atherosclerosis in the setting of SLE is yet to be fully elucidated. High-density lipoprotein was reported to be signifi cantly dysfunctional and proinfl ammatory in SLE patients, and correlated with measures of subclinical atherosclerosis [52]. Endothelial cell dysfunction in SLE leads to abnormal vascular reactivity and repair, contributing to the accelerated atherogenesis [53]. Interferon activity was independently associated with subclinical measures of atherosclerosis in a cohort study [54]. Specifi c subtypes of peripheral blood mononuclear cells, the low-density granulocytes, and increased Toll-like receptor signaling have been proposed to contribute to higher interferon production and vascular dysfunction in patients with SLE [55,56]. Autoantibodies including antiphospholipid and antilipoprotein antibodies have been associated with abnormal vascular function and atherosclerosis in SLE [57].
Despite the evidence of a link between infl ammation and atherosclerosis in SLE, multiple studies have failed to show any consistent association of coronary calcium scores or carotid plaques with markers of disease activity [51,58,59]. In patients with SLE, the proportion of noncalcifi ed vulnerable plaque was shown to be increased compared with the calcifi ed stable plaque, and correlated with measures of disease activity [60]. Elevated homocysteine, asymmetric dimethylarginine, leptin and highsensitivity C-reactive protein levels have been proposed as markers of accelerated atherosclerosis in SLE in some studies [61][62][63][64].
Th e management of atherosclerosis in SLE is currently limited to the control of traditional risk factors. Statins are believed to have anti-infl ammatory properties, in addition to their lipid-lowering eff ects. However, two large randomized controlled trials failed to show any benefi cial eff ects of 20 mg atorvastatin versus placebo in adult and pediatric SLE patients without clinical cardiovascular disease [65,66]. Th e use of statins in SLE patients should thus be limited to treat hyperlipidemia. Hydroxychloroquine was noted to have weak protective eff ects on cardiovascular risk in SLE [28]. Mycophenolate reduced the atherosclerotic burden in mice models but failed to reduce progression of subclinical atherosclerosis in a large cohort study [67,68]. Several trials evaluating anti-interferon therapy in SLE are currently ongoing, and benefi cial eff ects on athero sclerosis might become evident. However, no SLE treatment is currently of proven benefi t to reduce the risk of or halt the progression of atherosclerosis in SLE.

Corticosteroid-associated damage in systemic lupus erythematosus
Corticosteroids are the mainstay of therapy for SLE, with proven effi cacy. Th e harm they cause in the short term and the long term, however, is one of the major issues in SLE management. Corticosteroids increase the traditional cardiovascular risk factors, including serum lipids, blood pressure, weight and glucose. Fifteen years after the diagnosis of SLE, the majority of permanent organ damage can be attributed to the corticosteroids. Although the risk of damage is higher with high doses, there is no safe dose for chronic use. Even small doses, if continued long enough, will signifi cantly increase the morbidity [69]. Th e cumulative dose of corticosteroids has signifi cant association with cataracts and osteo porotic fractures. Both the current dose and the cumulative dose are associated with cardiovascular events [70]. When adjusted for confounding by indication due to SLE disease activity, the hazard ratio for organ damage increases dramatically with prednisone doses of 6 to 12 mg/day (hazard ratio = 1.5), 12 to 18 mg/day (hazard ratio = 1.64) and >18 mg/day (hazard ratio = 2.51) [69]. Every attempt should be made to minimize the dose and duration of corticosteroid exposure.

Quality of life in systemic lupus erythematosus
Th e QoL in patients with SLE is signifi cantly lower than in healthy controls and patients with other chronic diseases [71,72]. Disease activity and organ damage did not consistently correlate with lower QoL in SLE. Instead, the major predictors of poor QoL in SLE are nondiseasespecifi c variables including fatigue, chronic pain and mood disturbances [71,72].
Fatigue is a common and often crippling symptom experienced by about 85 to 92% of patients with SLE, with 50% rating it as the most disabling symptom [73]. Fatigue can signifi cantly contribute to the poor QoL in SLE patients [72]. Th e pathophysiological mechanisms of SLE-related fatigue are probably multifactorial, with a predominant role being played by the psychological domains. Psychosocial factors such as mood disorders, anxiety, poor sleep quality and chronic pain syndrome have shown consistent associations with fatigue in SLE [74,75]. Exercise programs have been shown to have a positive impact on fatigue among SLE patients [76]. However, fatigue remains an unmet need, refl ected by the fi nding that 81% of SLE patients feel that the healthcare service did not support them enough in the management of SLE-related fatigue [77].
Fibromyalgia is a chronic pain disorder characterized by widespread generalized pain, often associated with fatigue, anxiety and sleep disturbances. Th e prevalence of fi bromyalgia is much higher in SLE patients, compared with the general population [78]. Fibromyalgia in SLE contributes to the lower QoL and correlates with psychosomatic and aff ective variables but not with disease activity or damage [78][79][80]. Th e widespread pain of concomitant fi bromyalgia can lead to diagnostic confusions and potential overtreatment if symptoms are mistaken for SLE disease activity.
Mood disturbances are very common in patients with SLE, with depression being the most prevalent aff ective symptom [81,82]. Depression contributes to the fatigue and cognitive dysfunction, and signifi cantly correlates with lower QoL in patients with SLE [45,71,83]. Although psychological eff ects of dealing with a chronic disease may contribute to the high prevalence of depression, disease-specifi c mechanisms probably play a role. Associa tions with specifi c antibodies and alterations in cerebral blood fl ow have been reported in depressed SLE patients [84,85]. However, the data are not conclusive and depression in patients with SLE should be treated with conventional measures similar to the general population.

Perspective on future therapeutics and clinical research in SLE
Despite advances in therapies, the control of disease activity in SLE remains suboptimal. Flares are common and sustained disease control is generally limited to a small fraction of patients [86]. Th ese fi ndings suggest that, despite signifi cant improvements in SLE treatment, conventional approaches have probably reached their maximal benefi t and alternate options have to be considered. Multiple new agents with immuno modulatory eff ects have been investigated in recent years but limited success has been achieved. Two large randomized controlled trials evaluating rituximab for treatment of SLE failed to meet their primary endpoints, despite good effi cacy data in open-label studies [23,87]. Only belimumab showed effi cacy in randomized controlled trials and received US Food and Drug Administration approval for treatment of SLE [88,89]. Th ese studies have raised important issues that should be addressed in future SLE research.
Th e heterogeneous nature of the disease makes it diffi cult to design clinical studies in SLE. In the absence of specifi c biomarkers, classifi cation criteria are generally used to defi ne study populations. Although these criteria encompass the breadth of the disease spectrum, patients with diverse manifestations and probably diff erent pathogenic mechanisms will be grouped together. Th is limitation can be avoided by defi ning specifi c disease subgroups based on organ manifestations, such as renal disease. However, this would seriously limit the eligible patient population, however, stressing the need for multicenter collaborative projects. Limiting the inclusion in this manner may not be even a viable option if uncommon manifestations are considered. Currently, some degree of heterogeneity in study popula tions has to be accepted.
Another major issue in SLE research has been the choice of outcome measures. Th e US Food and Drug Administration draft guidance statement on SLE clinical trials suggested the use of disease activity indices to measure the effi cacy of the intended intervention [90]. Several such indices have been developed and validated for use in clinical trials. Some disease activity indicessuch as the SLE Disease Activity Index (SLEDAI) and its variants (SLEDAI-2K, SELENA-SLEDAI) -measure overall disease activity, while others -such as the British Isles Lupus Assessment Group Index (BILAG) -measure organ-specifi c activity. Physician's global assessment uses the treating physician's overall assessment to assign a numerical value to the disease activity on a visual analog scale of 0 to 3 [91]. Th is assessment has been used to defi ne fl ares (>1 point rise) in clinical trials [91]. Physician visual analog scale rating was found to have high variability in a comparative study of outcome measures by the SLICC study group [92], but it has been successfully incorporated as part of the SLE Responder Index (SRI) [93]. Th e SLE Responder Index is a composite index developed to incorporate the strengths of diff erent disease activity indices. Th is index provides a comprehensive defi nition of meaningful clinical response and has been used to defi ne the primary end point in clinical trials [93]. Th e SLE Responder Index utilizes the SELENA-SLEDAI score to determine global improvement, British Isles Lupus Assess ment Group Index domain scores to ensure no signifi cant worsening in previously unaff ected organ systems, and physician's global assessment to ensure that improvements in disease activity are not achieved at the expense of the patient's overall condition [93]. Th e SLE Responder Index has been used successfully in the belimumab phase 3 trials and has the potential to serve as an outcome measure in the future SLE therapeutic trials.
Th e background immunosuppression used as the standard of care in SLE trials adds another confounder to the picture. Most studies have employed diverse background treatments in both the treatment and placebo groups, to which the candidate agent is added. In addition, treatment adjustments have been either mandated or at least allowed during the studies. However, these concomitant therapies may have their own eff ects, masking the effi cacy of the target intervention. Examples include the failed LUNAR and EXPLORER trials, in which rituximab or placebo were added to background therapy including high-dose corticosteroids and immunosuppressive drugs such as mycophenolate [23,87]. Effi cacy of rituximab over placebo was not found in these trials, despite a large body of evidence favoring rituximab in observational and cohort studies. In contrast, the belimumab trial design permitted early tapering of cortico steroids and may have contributed, at least partially, to the positive results [88,89]. However, designing a trial in SLE where the active arm receives only the experimental agent will be diffi cult and unethical. Th ere is a need to develop a clearly defi ned standard of care control arm, against which the newer agents can be tested.
Th e fi eld of SLE clinical and therapeutic research is advancing rapidly. Large international collaborations have resulted in development of new criteria, composite outcome indices and insights into disease pathogenesis. Multiple newer biologic agents targeting specifi c immune system pathways and eff ectors are undergoing evaluation [94]. Hopefully, these eff orts will lead to development of newer therapeutic agents in SLE, a dire need.

Conclusions
Th e management of SLE remains a challenge despite signifi cant advances in the treatment. Th e new SLICC classifi cation criteria with better sensitivity will probably help in better identifi cation of patients, but we have not yet reached the stage where early diagnosis can be universally achieved. Th e fi eld of disease biomarkers is rapidly evolving with many putative candidates. However, no biomarker has been successfully validated for use in the routine clinical setting. Instead of a single measure, a battery of markers may perhaps be developed in the future to predict diff erent disease aspects in a heterogeneous disease such as SLE. Refractory disease manifestations and development of damage associated with disease and therapies remain outstanding issues. Premature mortality and higher morbidity from atherosclerosis in this pre dominantly young group of patients are a major concern. Fatigue is one of the most prevalent and disabling symptom in SLE, signifi cantly contributing to the poor QoL. Th e quest for newer targeted therapies has met with limited success despite many clinical trials. Whether there is a true lack of effi cacy or whether the trial designs in SLE are partly to blame is open to discussion. However, advances have been made in developing reliable outcome measures for clinical research. Future research will focus on the goals of increasing survival, limiting organ damage and improving QoL for patients with SLE.