Introduction

Systemic lupus erythematosus (SLE) is a chronic, systemic, autoimmune inflammatory disease of unknown aetiology,1 and is characterized by autoantibody production and protean clinical manifestations. The disease incidence has a striking 9:1 female predominance and peaks during childbearing years.1 The overall prevalence of SLE is 1:2,000, and it is more common in particular ethnic or racial groups, such as Asian and Afro-Caribbean people.1 Although arthritis and skin disease are the most frequent manifestations of SLE, visceral involvement is more serious. In particular, involvement of the lungs, kidneys and central nervous system accounts for most of the morbidity and mortality attributed to SLE.2,3,4 Of these visceral manifestations, nervous system involvement poses the greatest clinical challenge.

Nervous system involvement in SLE encompasses a variety of neurological and psychiatric features. Neuropsychiatric events attributed to SLE (NPSLE) are primary manifestations of the disease, rather than complications of the disease (for example, hypertension) or its therapy (for example, infection), or concurrent, non-SLE neuropsychiatric disease (Figure 1). The prevalence of NPSLE, which ranges from 21% to 95%,5,6,7,8,9,10,11,12 and the prognosis following a neuropsychiatric event are both highly variable. Several studies of NPSLE have reported an increase in mortality,13,14,15,16 whereas others have not.17,18,19 This variance could reflect differences in research methodology, bias in the selection of patients for study, or changes over time in global disease severity. Many early studies of NPSLE prognosis were limited by several factors: lack of standardized definitions or criteria for neuropsychiatric events, including the attribution of neuropsychiatric events to SLE and non-SLE causes; failure to use validated instruments to measure important outcomes such as organ damage and quality of life; and retrospective, single-centre study design.

Figure 1: Factors contributing to neuropsychiatric events in patients with SLE.
figure 1

Focal and diffuse nervous system events can result from autoimmune or inflammatory mechanisms directly related to SLE (primary NPSLE), as a consequence of complications of the disease (for example, uraemia or hypertension) or its therapy (for example, infection) (secondary NPSLE), or as a concurrent neuropsychiatric event unrelated to SLE. Abbreviations: +, minor contribution to NPSLE; ++, moderate contribution to NPSLE; +++, substantial contribution to NPSLE; aPL, antiphospholipid antibodies; NPSLE, neuropsychiatric SLE; SLE, systemic lupus erythematosus. Adapted with permission from Springer © Hanly, J. G. Curr. Rheumatol. Rep. 3, 205–212 (2001).

PowerPoint slide

Full size image

This article reviews advances in this important aspect of SLE. Current thinking on the classification and attribution of neuropsychiatric events in patients with SLE is discussed. Pertinent aspects of immunopathogenetic pathways are reviewed, and provide the basis for a proposed diagnostic and therapeutic approach to NPSLE.

Clinical manifestations of NPSLE

Classification

Many classifications of NPSLE lack definitions of individual manifestations and standardization for investigation and diagnosis. In 1999, the ACR produced a standard nomenclature and set of case definitions for 19 neuropsychiatric syndromes known to occur in SLE (as listed in Box 1).20 These syndromes can be segregated into central and peripheral,20 diffuse and focal neuropsychiatric events.21 The ACR classification is comprehensive in the scope of neuropsychiatric manifestations it describes, and provides guidance on investigations and diagnostic criteria for each. However, the classification is not, and in fact was never intended to be, specific for neuropsychiatric events caused exclusively by SLE. Thus, whether using the ACR classification in clinical practice or as part of a research study, it is important to attribute events to SLE and non-SLE causes to optimize the care of individual patients presenting with neuropsychiatric events or to ensure the validity of clinical research studies of NPSLE. To this end, the ACR classification lists causes other than SLE that could be responsible, in part or entirely, for each of the neuropsychiatric syndromes. This component of the ACR classification, in combination with other variables such as the relationship between time of onset of neuropsychiatric events to diagnosis of SLE and the high frequency of some neuropsychiatric events in the general population, have been used to develop attribution models for neuropsychiatric events in SLE.22,23

Depending upon the stringency of the attribution models, neuropsychiatric events occur in 6–12% of patients with newly diagnosed SLE over the first year of the illness, and the proportion attributed to SLE varies from 19% to 38%.22 Although headache and mood disorders are the most frequent neuropsychiatric complaints overall in patients with SLE, seizure disorders, cerebrovascular disease, acute confusional states and neuropathies are the most common neuropsychiatric syndromes attributed to SLE. The cumulative occurrence of neuropsychiatric events increases over time, although the proportion of events attributed to SLE and non-SLE causes remains the same.24,25

Regardless of attribution, neuropsychiatric events in patients with SLE are associated with a substantial negative effect on health-related quality of life (HRQoL) (Figure 2), even when factors such as global SLE disease activity, cumulative organ damage and medications are taken into account.22,23 Correctly attributing neuropsychiatric events to SLE and non-SLE causes is critical to determining the most appropriate treatment plan.

Figure 2: Negative effect of NP events on HRQoL in patients with SLE.22
figure 2

The difference in HRQoL of patients newly diagnosed with SLE with and without NP events is indicated by SF-36 subscale scores (upper panel) and physical and mental composite scores (lower panel), expressed as means ±SEM. Those patients with NP events, regardless of attribution to SLE or non-SLE causes, had consistently lower scores, indicating poorer HRQoL. Abbreviations: HRQoL, health-related quality of life; NP, neuropsychiatric; SF-36, short form 36; SLE, systemic lupus erythematosus. Reproduced with permission from John Wiley & Sons, Inc. © Hanly, J. G. et al. Arthritis Rheum. 56, 265–273 (2007).22

PowerPoint slide

Full size image

Cognitive dysfunction in SLE

Cognition is the sum of intellectual functions that result in thought, and includes reception of external stimuli, information processing, learning, storage and expression. Disturbance of even one of these functions can result in disruption of normal thought production and present as cognitive dysfunction. Cognitive complaints and objectively confirmed cognitive impairment are frequently encountered in patients with SLE. Although cognitive impairment can be viewed as a distinct syndrome, it can also serve as an indicator of overall brain health, which can be affected by a number of factors including other neuropsychiatric syndromes. Owing to the poor correlation between cognitive symptoms and objective findings on formal neuropsychological assessment,26 the presence, characteristics and severity of cognitive impairment should first be confirmed by formal neuropsychological assessment. Cognitive dysfunction identified in this way has been reported in up to 80% of patients with SLE,27 although most studies report a prevalence in the range 17–66%.28,29 Many individual patients have subclinical cognitive deficits. For example, a review of 14 cross-sectional studies of cognitive function in SLE revealed subclinical cognitive impairment in 11–54% of patients.28 A single pattern of SLE-associated cognitive dysfunction has not been found, but commonly identified cognitive abnormalities include overall cognitive slowing, decreased attention, impaired working memory and executive dysfunction (for example, difficulty with multitasking, organization or planning).

Aetiology and pathogenesis of NPSLE

Neuropsychiatric events attributable to SLE—that is, as primary manifestations of the disease—are thought to arise from vascular abnormalities, autoantibodies and inflammatory mediators. As discussed below, these factors interact in various pathogenic mechanisms of NPSLE.

Contributing factors

Vascular abnormalities

A bland, noninflammatory microangiopathy in association with brain microinfarction is the predominant neuropathological finding of NPSLE.30,31 By contrast, inflammatory disease of small or large blood vessels (vasculitis) is rare.

Autoantibodies

Antineuronal antibodies in NPSLE have been demonstrated to have a temporal relationship with neuropsychiatric events,32 to be present in cerebrospinal fluid (CSF)33 and to occur in postmortem neuronal tissues from patients with NPSLE.34 The occurrence of autoantibodies in the CSF is due to passive transfer from the circulation through increased permeability of the blood–brain barrier35 and, independently, to intrathecal production.33,35 A subset of anti-DNA antibodies in human SLE and in a mouse model of the disease crossreact with neuronal NR2 glutamate receptors.36 In contrast to earlier studies of antineuronal antibodies, anti-NR2 antibodies induce apoptotic cell death in vitro36 and in vivo.37 In animal models, enhanced permeability of the blood–brain barrier is critical for the access of anti-NR2 antibodies to neuronal cells.38 The blood–brain barrier can be permeabilized by both SLE factors (for example, immune complex deposition or cytokines) and non-SLE factors (for example, smoking or hypertension). In human SLE, the association between circulating anti-NR2 antibodies and NPSLE is inconsistent,39 but the strongest association has been found with autoantibodies in the CSF.40,41 Anti-ribosomal P antibodies have been associated with NPSLE, particularly psychosis, in some but not all studies.42,43,44,45,46,47 Discrepancies between these studies might be attributable to differences in diagnostic criteria for psychiatric disease, variance in the temporal relationship between clinical events and serological testing, and differences in assay technique (perhaps related to antigen preparation and purity).

Autoimmune antiphospholipid antibodies (aPL), directed against phospholipid-binding proteins such as β2-glycoprotein I and prothrombin, induce a procoagulant state48,49 and are associated predominately with focal manifestations of NPSLE. Most of these are vascular events, such as stroke,50 and seizure disorders.8 Additional studies have reported an association of aPL with cognitive impairment, even in the absence of stroke.50,51 The favoured pathogenetic mechanism for this family of autoantibodies in NPSLE is thrombosis within vessels of different calibres and consequent cerebral ischaemia. However, a possible direct pathogenic effect of aPL on neuronal cells is suggested by their intrathecal production in patients with NPSLE,35 their association with diffuse cognitive impairment,51,52 and their modulation of neuronal cell function in vitro.53

Inflammatory mediators

Proinflammatory cytokines might have a role in NPSLE. Initial studies reported associations between increased intracranial production of IL-6 with seizures,54 and the production of IFN-α with lupus psychosis.55 Subsequent studies provided further evidence of intrathecal production of IL-656,57,58 and other cytokines including IL-10,59 IL-260 and IL-858 in patients with NPSLE. Cytokines associated with NPSLE are produced by neuronal55,57 and glial cells,55 probably in response to autoantibodies within the intrathecal space.61,62 In vitro studies suggest that cytokine production occurs as a consequence of binding of immune complexes (formed of autoantibodies and RNA-protein antigens) to FcγRII on plasmacytoid dendritic cells, followed by endocytosis and activation of Toll-like receptor 7 (TLR7). Further support for this mechanism includes evidence of neuronal and glial degradation products (a potential source of antigen) in the CSF of patients with SLE,63 and findings of elevated CSF levels of matrix metalloproteinase 9,63 which increases the permeability of the blood–brain barrier, thus providing intrathecal access to circulating autoantibodies.

Pathogenic mechanisms

Taken together, the evidence presented earlier suggests two separate and potentially complementary autoimmune pathogenic mechanisms for NPSLE (Figure 3). The first mechanism implicates injury to large and small blood vessels, mediated by aPL, immune complexes and leukoagglutination. Clinical sequelae to this vascular-injury mechanism include focal neuropsychiatric events such as stroke and diffuse neuropsychiatric events such as cognitive dysfunction. The second mechanism involves autoimmune inflammation injury, with increased permeability of the blood–brain barrier, intrathecal formation of immune complexes, and production of IFN-α and other inflammatory mediators. Clinical sequelae of this mechanism include diffuse neuropsychiatric manifestations such as psychosis and acute confusional state.

Figure 3: Autoimmune pathogenesis of NPSLE.
figure 3

Vascular injury involving both large and small-calibre vessels is mediated by aPL, immune complexes and leukoagglutination, and can result in focal neuropsychiatric events (for example, stroke) or diffuse neuropsychiatric events (for example, cognitive dysfunction). Inflammatory injury, involving increased permeability of the BBB, formation of immune complexes and production of IFN-α and other inflammatory mediators, can lead to diffuse neuropsychiatric manifestations (for example, psychosis, acute confusional states). Abbreviations: aPL, antiphospholipid antibodies; BBB, blood–brain barrier; MMP, matrix metalloproteinase; NPSLE, neuropsychiatric systemic lupus erythematosus; pDC, plasmacytoid dendritic cell.

PowerPoint slide

Full size image

Diagnostic approach and investigation

In the assessment of a patient with SLE and a neuropsychiatric event, it is important to first determine if the neuropsychiatric event is a primary manifestation of the disease, a complication of the disease or its therapy, a coincidental disease process, or a combination thereof (Figure 1). Given the absence of a diagnostic gold standard for most neuropsychiatric events, the correct attribution is made largely by a process of exclusion. Information from a variety of sources might be utilized to a varying extent, depending upon the clinical circumstances (Box 2).

Generic investigations

As emphasized by the EULAR task force recommendations for the management of NPSLE,64 in patients with new or unexplained symptoms or signs suggestive of neuropsychiatric disease, the diagnostic work-up should consider any investigations that would be done in non-SLE patients presenting with the same manifestations. For example, a patient with SLE who presents with a transient ischaemic attack or stroke should undergo a screening echocardiogram and Doppler ultrasonography of the carotid arteries, in addition to testing for SLE-specific causes such as aPL. In addition, given the higher-than-expected frequency of premature cardiovascular disease in SLE patients,65,66 cardiac risk factors such as poorly controlled hypertension and hyperlipidaemia should also be considered.67 Similarly, investigations in patients who present with acute confusion should include the identification of metabolic abnormalities, infection and use of psychoactive drugs, and a search for diabetes mellitus and vitamin B12 deficiency should be undertaken in those who present with peripheral sensory neuropathy.

SLE disease activity

The evaluation of SLE disease activity in organ systems other than the nervous system is important to ensure appropriate management of neuropsychiatric events in patients with SLE. Such assessment might also help attribute these events to SLE and non-SLE causes. Some studies,67,68,69 but not all,70,71 have found an association between increased global SLE disease activity and neuropsychiatric events attributed to SLE. This association is probably more robust for diffuse rather than focal neuropsychiatric events.

Autoantibodies in the circulation and CSF

Of the circulating autoantibodies, aPL are likely to provide the greatest diagnostic yield, especially in patients with focal neuropsychiatric events or cognitive decline.50,51,52 The association between circulating anti-NR2 antibodies and NPSLE is less clear. Omdal et al.72 found an association between levels of these antibodies and depression, decreased short-term memory and learning. Apart from another report of an association of anti-NR2 antibodies with depression,73 three cross-sectional studies74,75,76 (one of which had a 5-year follow-up period74) did not confirm these findings. Yoshio et al.40 studied both serum and CSF samples from 80 patients with SLE (53 of whom had NPSLE) and found the strongest association between neuropsychiatric events and CSF autoantibodies. Another study of CSF samples from 56 patients with SLE found anti-NR2 antibodies in 44% and 82% of patients with focal and diffuse NPSLE, respectively.41 Thus, although in practice the examination of CSF is done primarily to exclude infection, its availability provides an opportunity to measure levels of this interesting autoantibody, which might contribute to the diagnosis of NPSLE in individual patients. The measurement of CSF cytokines and biomarkers of neurological damage holds considerable academic interest for the study of pathogenetic mechanisms, but at this time is neither recommended nor even feasible in practice.

Neuromyelitis optica (NMO), also known as Devic syndrome, is a severe demyelinating disorder of the central nervous system that causes longitudinal transverse myelitis of at least three vertebral segments and recurrent optic neuritis. NMO has been reported in patients with SLE77 and is associated with NMO-specific autoantibodies whose antigenic target is aquaporin 4,78 the most abundant water channel in the central nervous system.79 Although NMO is a rare clinical presentation, suspicion of this syndrome in a patient with SLE warrants the measurement of autoantibodies to aquaporin 4.

Electrophysiological studies

Electroencephalography is primarily used to investigate seizure disorders, and detects abnormalities including asymmetry of the electric cerebral activity, diffuse disorganized background activity and focal epileptiform discharges.80 Electrophysiological abnormalities in a study of 1,533 patients with SLE and various peripheral neuropathies in 207 (14%) patients, 60% of which were attributed to SLE, found evidence of axonal neuropathy in 70% and signs of demyelination in 20%.68

Neuropsychological assessment

Formal neuropsychological testing is carried out when there is a clinical suspicion of impaired cognitive ability. This formal testing does not need to be performed routinely in all patients, as the detection of isolated subtle subclinical cognitive deficits is not clinically important. At present, no simple screening test for cognitive dysfunction is available for use in patients with SLE, as most such tests lack sensitivity for mild but clinically relevant dysfunction. Self-report instruments used to screen for cognitive difficulties have been validated by some studies81,82 but not others.26 Computerized testing facilitates efficient screening of patients with SLE by non-experts,83 but formal testing remains the only definitive way to diagnose cognitive impairment; thus, patients with suspected impaired cognitive ability should be referred for full neuropsychological assessment. The battery of neuropsychological tests proposed by the ACR for the assessment of cognitive function in SLE20 is comprehensive, but its widespread use remains limited, as such test series are time-consuming, require specialized training and are subject to practice effects with repeated use. If abnormalities are detected by use of these tests, repeat testing using the same battery should be done after a reasonable interval, usually several months, to measure change in cognition following observation or treatment.

Neuroimaging

Standard and advanced neuroimaging of brain structure and function help to localize intracranial abnormalities, determine whether the lesions involve white or grey matter, and assess their chronicity and change over time. The assessment of brain structure by CT has largely been replaced by MRI, which is more sensitive (especially T2-weighted images). Abnormalities in patients with SLE include changes in white and grey matter, in addition to global and regional cerebral atrophy.84 However, with the exception of large cerebral infarcts, the correlation between structural changes and clinical neuropsychiatric manifestations of SLE is low, thus making MRI-determined structure a poor marker of disease progression or treatment outcomes.85 More advanced MRI methodology for the detection of structural abnormalities include magnetization transfer imaging (MTI), diffusion-weighted imaging (DWI) and diffusion-tensor imaging (DTI).86

MTI measures magnetization transfer between bound and unbound hydrogen molecules (for example, between white matter and CSF). Magnetization transfer is diminished by either a decrease in bound molecules (as occurs in demyelination) or an increase in unbound molecules (as in oedema). Decreased whole-brain magnetization transfer has been reported in patients with SLE, even in the absence of other MRI-detected structural changes,87 and tends to be greater in longstanding NPSLE compared with active or acute disease.86 Diffusion MRI enables the study of diffusion of water in the brain, which is dependent upon the interaction of water with macromolecules and membranes, thereby enabling the assessment of tissue architecture. DWI is particularly effective in identifying hyperacute brain injury, such as the acute ischaemia that follows a stroke, when the acute shift of fluid into the intracellular compartment and cytotoxic oedema restrict the diffusion of water.85,88 DTI uses a similar technology to assess the integrity of neural white matter tracts in the brain.86

PET is the most objective method of imaging brain function; however, its clinical utility is limited as is not widely available.89 Single-photon emission CT (SPECT)89 is used to analyse regional cerebral blood flow and metabolism, and is exquisitely sensitive in the detection of neuroimaging abnormalities. In patients with SLE,90,91,92,93,94 SPECT has identified both diffuse and focal deficits, which can be fixed or reversible. However, these findings are not specific for SLE90 and do not always correlate with clinical neuropsychiatric manifestations.95 In fact, findings of SPECT imaging can be abnormal in up to half of patients with SLE and no clinical manifestations of neuropsychiatric disease.84 Thus, the clinical relevance of these imaging abnormalities is not always clear.

Magnetic resonance angiography (MRA) enables the noninvasive visualization of cerebral blood flow. One potential drawback of this method is that it is not ideal for the visualization of blood flow in small-calibre vessels, which are the vessel type primarily involved in NPSLE. Magnetic resonance spectroscopy (MRS) measures biochemical compounds including N-acetylaspartate, choline and creatine within pre-determined regions of interest. Decreased levels of N-acetylaspartate, believed to reflect neuronal or axonal loss or dysfunction, have been reported in patients with SLE, even in the absence of visible damage on structural MRI scans.84,86 Whereas MRS examines biochemical changes in brain tissues, functional MRI (fMRI) measures changes in local brain deoxyhaemoglobin levels, which probably reflect neuronal activity and thereby provide an indirect measure of brain function. In a study of blood-oxygen-level-dependent fMRI (BOLD-fMRI),96 frontoparietal activation while performing a task that engaged working memory was greater in the nine patients with NPSLE than in the same number of patients with RA and healthy controls. These findings were thought to show an adaptation of neuronal function, through the recruitment of extra-cortical pathways, to compensate for the impaired function of standard pathways. Another study revealed differences in brain activation patterns between 10 patients with childhood-onset SLE and healthy controls.97 Thus, fMRI could be informative regarding dysfunction or redistribution of functions as a result of SLE.86

Despite its limitations, at this time, structural MRI remains the neuroimaging method routinely used in clinical practice. More advanced methods such as MTI, DTI, MRS and fMRI suffer from variability in results between scanners and a lack of normative or standardized guidelines for interpretation. In the future, however, multimodal imaging is likely to become the standard of practice in the diagnosis and monitoring of neurological disorders, including NPSLE.

Treatment of neuropsychiatric events

Given the variety of clinical manifestations and the challenges associated with diagnosis, attribution and treatment, the management of patients with SLE presenting with neuropsychiatric events is optimized by a multidisciplinary approach, ideally led by a rheumatologist (Box 3). As mentioned earlier, a thorough clinical assessment and appropriate investigations should be used to determine the attribution of the neuropsychiatric event. The identification and treatment of non-SLE-related factors is important in all cases, even in those patients in whom SLE is the main contributing factor. For example, any serious infections or metabolic abnormalities should be addressed in patients presenting with acute neuropsychiatric events such as acute confusion and seizures, and patients with vascular neuropsychiatric events should be screened for cardiovascular risk factors. Likewise, the use of pharmacological therapies for relieving anxiety and depression, improving poor sleep hygiene and maintaining normal blood pressure could relieve cognitive complaints. More specific therapies to address primary manifestations of SLE are selected on the basis of which immunopathogenic mechanism of NPSLE, namely inflammation-mediated injury or vascular-mediated injury, is predominant. Only one randomized controlled trial of immunosuppressive therapy in NPSLE has been undertaken,98 as highlighted by a 2013 Cochrane review of this therapy.99

Autoimmune-mediated inflammatory injury

The standard of care for patients with SLE and serious visceral involvement, in particular lupus nephritis, includes treatment with high-dose corticosteroids, azathioprine, cyclophosphamide and mycophenolate mofetil. A similar approach is inferred to be appropriate for those neuropsychiatric manifestations arising from autoimmune-induced inflammation, despite a relative lack of clinical evidence. In an open-label study of 13 patients with lupus psychosis, treatment with oral cyclophosphamide for 6 months followed by maintenance therapy with azathioprine produced clinical improvement.100 A 2-year randomized, controlled trial by Barile-Fabris et al.98 in 32 patients with acute, severe NPSLE reported a significantly better response to therapy with intermittent intravenous cyclophosphamide than with intravenous methylprednisone (95% versus 54%, P <0.03). More targeted immunosuppressive therapies, such as B-lymphocyte depletion with anti-CD20 antibody used alone or in combination with cyclophosphamide,101 are promising but require further study. In most studies in NPSLE, immunosuppressive therapy has been used in combination with corticosteroids, and alongside symptomatic therapies such as psychoactive medications.

Vascular injury

Treatment of focal neuropsychiatric disease attributed to aPL requires anticoagulation, and such therapy will usually be lifelong.27 In the absence of controlled clinical trials to guide the type and intensity of anticoagulation therapy required to treat NPSLE due to thrombosis, one must rely on studies of prevention of recurrent thrombosis (at any anatomical location) in patients with antiphospholipid syndrome. Two controlled studies in antiphospholipid syndrome found no significant difference between low-intensity treatment (target international normalised ratio [INR] 2.0–3.0) and high-intensity treatment (target INR >3.0) with warfarin in the prevention of recurrent thrombosis.102,103 However, a minority of patients in these studies had arterial thrombosis, and controversy remains on the optimal target INR for the management of such cases.104 Potential adjunctive therapies to consider, especially in patients with recurrent thrombosis whilst on warfarin, are antiplatelet agents, antimalarial agents105 and statins.106

Cognitive impairment

Non-SLE causes of cognitive dysfunction, including sleep deprivation, mood disorders, fatigue and medications, should be sought and addressed. Several medications commonly used for the treatment of SLE, such as antidepressant, anticonvulsant and antihypertensive agents, can induce reversible cognitive impairment; these effects might be improved by changes in the choice or dose of drug. Pharmacological treatment and cognitive rehabilitation can also be considered.

In a placebo-controlled trial of pharmacologic therapy for SLE-associated cognitive dysfunction in 10 patients with mild SLE, daily treatment with 0.5 mg/kg prednisone reportedly improved cognition in five of eight patients who completed the trial.107 In a single-centre, placebo-controlled study in 51 patients with SLE, memantine, an NMDA-receptor antagonist used to treat Alzheimer disease, did not improve cognitive performance compared with placebo over 12 weeks.108 Given the observed association between cognitive impairment and aPL,51,52 the use of antiplatelet or anticoagulant therapy in aPL-positive patients is logical; however, evidence for the efficacy of this approach is lacking. The use of immunosuppressive therapy in patients with presumed inflammation-mediated brain injury is also logical, but again lacks evidence of efficacy from controlled trials, especially in patients with cognitive impairment as the sole manifestation of NPSLE.

Cognitive rehabilitation therapy aims to help individuals functionally adapt to cognitive deficits, typically through intensive retraining of cognitive skills. This approach has been applied in numerous conditions, including stroke, dementia, traumatic brain injury and multiple sclerosis. One study of cognitive rehabilitation enrolled 17 women with SLE aged 25–60 years who reported cognitive difficulties that either interfered with adaptive functioning or caused emotional distress.109 All participants completed the MINDFULL (Mastering the Intellectual Navigation of Daily Functioning and Undoing the Limitations of Lupus) programme, which involved eight weekly 2 h “psychoeducational group intervention” sessions focused on cognitive-strategy training applied across multiple real-life situations. In addition, the MINDFULL programme included a psychosocial support component. The feasibility of this approach for the SLE population was demonstrated by the study's 100% retention rate. Patients reported better affect and overall quality of life, as well as memory self-efficacy (which has been linked to cognitive performance in daily life). Although these results are encouraging, controlled and long-term studies are required to confirm these preliminary findings and determine their durability over time.

Prognosis

The outcome of neuropsychiatric events in patients with SLE, whether attributed to SLE or non-SLE causes, has been examined in only a small number of studies. Clinical trials of therapy for these disorders have been uncontrolled, of short duration or focused upon a single neuropsychiatric manifestation.98,100,101,107,110,111 Longitudinal studies of cognitive function have generally reported stable performance on cognitive tests over time, with persistent or progressive cognitive dysfunction seen in only a minority of patients.112,113 Outcomes of observational cohorts have been inconsistent. For example, some studies have reported increased mortality in patients with neuropsychiatric events,13,14,15,16 whereas others have not.17,18,19 In a follow-up study of 32 patients hospitalized for NPSLE, neurological deficits had either substantially improved (69%) or stabilized (19%) after 2 years.117 In a prospective study of patients with SLE followed for up to 7 years (mean 3.6 years),24 approximately 15% of neuropsychiatric events were resolved at each annual assessment; however, most events persisted. Notably, whether or not an event resolved could not be predicted by its attribution to SLE or non-SLE causes. In contrast to this finding, two reports from a cohort study of patients recently diagnosed with SLE demonstrate a more favourable short-term outcome over a mean follow-up of 3.7 months23 and 1.9 years25 for neuropsychiatric events attributed to SLE compared with non-SLE events (Figure 4). The possibility exists that early treatment of NPSLE leads to better outcomes, akin to the therapeutic 'window of opportunity' seen in other rheumatic diseases.118,119

Figure 4: Physician-generated outcome scores for NP events at enrolment into an international inception cohort of patients with SLE.23
figure 4

NP events were attributed to SLE or non-SLE causes using one of two different attribution models. Using attribution model A, the NP event was attributed to SLE unless onset of NP events occurred before the study enrolment window, non-SLE factors that contributed to or were responsible for the NP event were identified, or NP events with a high frequency in the general population (as defined by Ainiala et al.5) occurred. Using attribution model B, the NP event was attributed to SLE unless onset of NP events occurred >10 years before the diagnosis of SLE, non-SLE factors responsible for the NP event were identified, or NP events with a high frequency in the general population (as defined by Ainiala et al.5) occurred. Those NP events that were attributed to SLE using either model A or model B had a significantly better outcome than NP events not attributed to SLE (P <0.001). Abbreviations: NP, neuropsychiatric; SLE, systemic lupus erythematosus. Reproduced with permission from John Wiley & Sons, Inc. © Hanly, J. G. et al. Arthritis Care Res. 59, 721–729 (2008).23

PowerPoint slide

Full size image

Conclusion

The occurrence of neuropsychiatric events in patients with SLE poses a diagnostic and therapeutic challenge. The precise characterization and correct attribution of these events to SLE and non-SLE causes is critical. The correct diagnosis relies heavily on careful clinical assessment of the patient and the selection of appropriate investigations. Recent studies have provided insight into the immunopathogenetic mechanisms of NPSLE, the contribution of non-SLE factors and the short-term and long-term prognosis of patients presenting with NPSLE. Current therapeutic strategies are largely empiric, based on known immunopathogenetic mechanisms and what has been observed from the treatment of other serious organ disease in SLE. Further insight into the immunopathogenetic mechanisms and clinical outcomes of NPSLE are required to inform the design and execution of therapeutic clinical trials.

Review criteria

The articles cited in this Review were selected from the author's personal library of articles on neuropsychiatric SLE. Selections were made on the basis of the expert opinion of the author. The reference list was last updated in October 2013.