The diagnosis and clinical management of the ...

Journal of Autoimmunity xxx (2016) 1e32

Contents lists available at ScienceDirect

Journal of Autoimmunity

journal homepage: locate/jautimm

Review article

The diagnosis and clinical management of the neuropsychiatric manifestations of lupus

M. Govoni a, *, A. Bortoluzzi a, M. Padovan a, E. Silvagni a, M. Borrelli b, F. Donelli b, S. Ceruti b, F. Trotta a

a Department of Medical Sciences, Rheumatology Unit, University of Ferrara and Azienda Ospedaliero-Universitaria Sant'Anna, Cona, Ferrara, Italy b Neuroradiology Unit, Department of Neuroscience and Rehabilitation, Azienda Ospedaliero-Universitaria Sant'Anna, Cona, Ferrara, Italy

article info

Article history: Received 19 June 2016 Accepted 21 June 2016 Available online xxx

Keywords: Systemic Lupus Erythematosus Neuropsychiatric lupus Central and peripheral nervous system Neuroimaging

abstract

Neuropsychiatric (NP) involvement in Systemic Lupus Erythematosus (SLE), can be a severe and troubling manifestation of the disease that heavily impacts patient's health, quality of life and disease outcome. It is one of the most complex expressions of SLE which can affect central, peripheral and autonomous nervous system. Complex interrelated pathogenetic mechanisms, including genetic factors, vasculopathy, vascular occlusion, neuroendocrine-immune imbalance, tissue and neuronal damage mediated by autoantibodies, inflammatory mediators, blood brain barrier dysfunction and direct neuronal cell death can be all involved. About NPSLE a number of issues are still matter of debate: from classification and burden of NPSLE to attribution and diagnosis. The role of neuroimaging and new methods of investigation still remain pivotal and rapidly evolving as well as is the increasing knowledge in the pathogenesis. Overall, two main pathogenetic pathways have been recognized yielding different clinical phenotypes: a predominant ischemic-vascular one involving large and small blood vessels, mediated by aPL, immune complexes and leuko-agglutination which it is manifested with more frequent focal NP clinical pictures and a predominantly inflammatory-neurotoxic one mediated by complement activation, increased permeability of the BBB, intrathecal migration of autoantibodies, local production of immune complexes and pro-inflammatory cytokines and other inflammatory mediators usually appearing as diffuse NP manifestations. In the attempt to depict a journey throughout NPSLE from diagnosis to a reasoned therapeutic approach, classification, epidemiology, attribution, risk factors, diagnostic challenges, neuroimaging techniques and pathogenesis will be considered in this narrative review based on the most relevant and recent published data.

? 2016 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 2. Classification (historical survey) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 3. Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 4. Attribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 5. Risk factors for NPSLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 6. Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 7. NPSLE pathogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00

7.1. Genetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 7.2. Vasculopathy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 7.3. Inflammatory mediators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 7.4. Autoantibodies (Abs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00

* Corresponding author. Department of Medical Science, Section of Rheumatology, University of Ferrara, Via Aldo Moro 8, 44124, Cona, Ferrara, Italy.

E-mail address: gvl@unife.it (M. Govoni).

0896-8411/? 2016 Elsevier Ltd. All rights reserved.

Please cite this article in press as: M. Govoni, et al., The diagnosis and clinical management of the neuropsychiatric manifestations of lupus, Journal of Autoimmunity (2016),

Downloaded from at GOT-CILEA-Anna - Ferrara Hospital Library July 18, 2016. For personal use only. No other uses without permission. Copyright ?2016. Elsevier Inc. All rights reserved.

2

M. Govoni et al. / Journal of Autoimmunity xxx (2016) 1e32

7.5. Bloodebrain barrier (BBB) dysfunction and cerebrospinal fluid autoantibodies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 8. Neuroimaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 9. Acute lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 10. Chronic lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 11. Advanced MRI techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 12. Nuclear Medicine techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 13. Therapeutic approach to NPSLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00

13.1. Mild neuropsychiatric lupus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 13.1.1. Headache . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 13.1.2. Mood disorder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 13.1.3. Cognitive dysfunction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00

13.2. Severe neuropsychiatric manifestations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 13.2.1. Cerebrovascular disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 13.2.2. Psychosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 13.2.3. Seizure disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 13.2.4. Myelopathy and demyelinating syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00

13.3. Management of less common manifestations: acute confusional state, movement disorder and aseptic meningitis . . . . . . . . . . . . . . . . . . . . . 00 13.4. Peripheral nervous system disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00

13.4.1. Cranial neuropathies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 13.4.2. Peripheral involvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 13.5. Treatment options in refractory NPSLE: rituximab, IVIG, plasma-exchange and autologous stem cells transplantation . . . . . . . . . . . . . . . . . . . 00 14. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00

1. Introduction

Neuropsychiatric (NP) involvement in Systemic Lupus Erythematosus (SLE), can be a severe and troubling manifestation of the disease that heavily impacts patient's health and disease outcome. It has now well accomplished that, in SLE patients, the occurrence of NP events is associated with a lower quality of life and poor prognosis [1] with a reported tenfold increase in mortality rate in NPSLE, compared with the general population [2].

NP involvement is one of the most complex expressions of SLE, deriving either from the anatomy and physiopathology of the nervous system itself and from the heterogeneity of clinical pictures, which can affect central (CNS), peripheral (PNS) and autonomous nervous system. Indeed, the phenomenology of NP involvement (NPSLE) can take a variety of aspects, NP events being focal or diffuse, mild or disabling, acute or chronic, active or not active, single or multiple, occurring contemporary or sequentially over time. Finally, given such heterogeneity, it appears quite intuitive that different pathogenetic pathways drive different clinical NP phenotypes.

In the attempt to depict a journey towards a reasonable approach to the treatment of NPSLE, classification, epidemiology, attribution, risk factors, diagnostic challenges, neuroimaging and more recent views on pathogenesis will be considered in this review.

2. Classification (historical survey)

The history of NPSLE classifcation begins a long time ago, in 1971, when seizures, psychosis and focal NP pictures were included in a first preliminary SLE classification [3]. Some years later, Kassan and Lockshin focused their attention on the complexity of NPSLE and its pathogenesis, highlighting the importance to consider confounding variables, the relevance of the chronological course of NP events and their functional impact [4]. In 1982 the ARA revised criteria for SLE formally included only seizures and psychosis [5]. In 1985 How et al. [6] proposed a classification model distinguishing MAJOR from MINOR NP events, suggesting that the satisfaction of one major criterion or, in the case of minor manifestations, the

additional presence of instrumental (EEG, brain scan) or lab (i.e. cerebro spinal fluid) abnormalities, should be considered to diagnose NPSLE. In this model, the relevance to rule out other possible aetiologies such as infection, drugs, uremia or hypertension was also emphasized, introducing the concept of "secondary NPSLE". In a consensus conference held in 1987, a group of international experts selected atypical psychosis, seizures, transverse myelitis and global cognitive dysfunction from a list of more than 50 possible clinical, laboratory and imaging abnormalities as part of NPSLE. The conclusions of this meeting were published in 1990 by Singer and Denburg [7].

In April 1997 an ad hoc multidisciplinary committee of 35 members, encompassing different areas of expertise, including rheumatologists, neurologists, psychiatrists, neuro-psychiatrists and haematologists convened on behalf of the American College of Rheumatology Research Committee. In 1999 the committee developed a standard nomenclature and set of case definitions for 19 NP syndromes (12 CNS and 7 PNS manifestations) deemed as occurring in SLE (Table 1) [8].

Undoubtedly ACR classification can be considered a milestone in the development of our knowledge on NPSLE providing definition, to formally describe each clinical NP syndrome; diagnostic criteria, to define the characteristic of each NP event; exclusion criteria, aimed to rule out NP event not directly related to SLE; associations, to consider concomitant or pre-existing comorbidities as potential confounding factors (i.e. hypertension, smoking, diabetes ...) and a set of recommendations to ascertain each NP event.

Although this classification has represented an important step forward in the description and categorization of NP events in SLE, providing a useful tool for patient selection in clinical studies, when applied in every day clinical practice its utility has appeared of limited value.

In fact, all the included NP syndromes are not specific for SLE, and some of them are very not specific being frequently observed in the general population. In order to increase the specificity of the ACR nomenclature, some Authors have suggested a revised version of the NP syndromes list, excluding some milder, less specific and more subjective NP syndromes such as headache, mild cognitive dysfunction and mood disorders along with peripheral neuropathy

Please cite this article in press as: M. Govoni, et al., The diagnosis and clinical management of the neuropsychiatric manifestations of lupus,

Journal of Autoimmunity (2016), Downloaded from at GOT-CILEA-Anna - Ferrara Hospital Library July 18, 2016.

For personal use only. No other uses without permission. Copyright ?2016. Elsevier Inc. All rights reserved.

M. Govoni et al. / Journal of Autoimmunity xxx (2016) 1e32

Table 1 Neuropsychiatric syndromes according to the 1999 American College of Rheumatology classification [8].

Central NPSLE

Peripheral NPSLE

Focal NPSLEa

Aseptic meningitis Cerebrovascular disease Demyelinating syndrome Headache Movement disorder Myelopathy Seizure disorders Acute confusional state Anxiety disorder Cognitive dysfunction Mood disorder Psychosis

Guillain-Barre syndrome Autonomic neuropathy Mononeuropathy Myasthenia gravis Cranial neuropathy Plexopathy Polyneuropathy

Cerebrovascular disease Seizures Movement disorder Myelopathy Cranial neuropathy Polyneuropathy Mononeuropathy Myasthenia gravis Autonomic neuropathy Plexopathy Guillain-Barre syndrome

a The distinction between focal or diffuse syndromes is reported according to 2010 EULAR recommendations [9].

3

Diffuse NPSLEa Aseptic meningitis Demyelinating syndrome Headache Acute confusional state Anxiety disorder Cognitive dysfunction Mood disorder Psychosis

without electrophysiological confirmation. By ruling out these socalled "minor" NP events, the specificity of the remaining NP pictures was reported to increase from 46 to 93% [10].

It should be pointed out that these clinical pictures are frequently observed in SLE and, sometimes, they really represent a NP manifestation of the disease so that it appears reductive excluding them from a formal definition of NPSLE. However, in these cases, a careful further rigorous process of attribution - not provided by the ACR classification - is mandatory in order to establish or rule out a possible link between the NP event with the underlying SLE.

Another critical point of the ACR classification is that a number of other neurological clinical syndrome (i.e. neuromyelitis optica, posterior reversible encephalopathy syndrome known as PRES, pure sensory small fiber neuropathy) although increasingly observed in SLE, have not been included [11,12].

Acknowledging the 1999 ACR case definition for NPSLE, the revised Systemic Lupus International Collaborating Clinics (SLICC) classification criteria for SLE extended the number of NP pictures deemed as part of NP involvement of the disease, including mononeuritis multiplex, myelitis, peripheral or cranial neuropathy, and acute confusional state [13].

3. Epidemiology

A number of critical points make it difficult to provide a clearcut evaluation of both prevalence and incidence of NP involvement in SLE. Variability in clinical presentation, different stringency of selection criteria and heterogeneity in population studied, have represented common obstacles.

In keeping with these limitations the reported overall prevalence of NP disease ranges widely between 14 to about 95% [10e26] so it is very difficult to establish what is the real burden of NPSLE. Interestingly this wide range has not significatively changed after the introduction of the ACR nomenclature if compared to that previously reported, with frequencies ranging from 37 to 95% [20e24]. However, considering some of the largest SLE cohorts (including several hundreds of patients) published after the 1999, the prevalence of NPSLE appears lower, ranging from 19 to 57%) [23,27e30]. Also bear in mind that different NP syndromes have a different prevalence and incidence, being those less specific (headache, cognitive dysfunction, mood disorders ...) the most frequently observed too.

On this background, from a clinical point of view, a more useful and pragmatic approach, is to stratify the frequency of specific NP events (now included in the ACR 1999 classification) as reported in Table 2.

Besides the definition and classification criteria used, two of the major determinants which make the difference in the assessment

of the overall prevalence of NP syndromes in SLE are the inclusion or exclusion of minor and nonspecific NP syndromes and e especially e the inclusion of cognitive dysfunction (from mild to severe). In a recent multicentric lupus Canadian cohort including 1253 SLE patients, with mean disease duration of 12 years, when using the most stringent ACR criteria definition (comprising only seizures and psychosis) a 6.4% prevalence was found, whilst when less restrictive criteria including minor nonspecific NPSLE manifestations was applied, the prevalence raised up to 38.6% [37].

In a 3 year prospective study of 370 SLE patients without previous history of CNS involvement, after excluding non-specific minor CNS complaints and PNS syndromes, the resulting prevalence of major CNS events among SLE patients was 4.3% with an estimated incidence rate of 7.8 events/100 person-years [38]. Again, when cognitive dysfunctions are screened by a structured battery test, the frequency of NP involvement goes up to 57e80% [20,23].

In a recent meta-analysis, after pooling all available studies for a total amount of 5057 SLE patients, Unterman et al. reported a prevalence of NPSLE of 44.5% in prospective studies and 17.6% in the retrospective ones. In a subanalysis of the 10 higher quality prospective studies (2049 patients) the overall prevalence of NP syndromes in SLE patients was estimated to be 56.3%, predominantly affecting CNS (93.1%) rather than PNS (6.9%) [39]. In summary, a realistic estimate of the overall prevalence of NP involvement in SLE suggests that nearly half of SLE patients will develop NPSLE during their disease course thus pointing out that NP involvement as a whole, is not a rare evenience and needs to be carefully checked for.

It is well acknowledged that SLE does occur more frequently in females of child-bearing age [40] whereas a similar sex-related effect is not so evident in NPSLE; on this topic literature data are conflicting and correction for confounders such as ethnicity, duration of follow-up, age of onset and comorbidites has rarely been assessed [41].

More NP involvement in men, such as peripheral neuropathy [42], seizures [43,44], cerebritis [45] and stroke [46], have been reported by some authors, while some others observed a decreased prevalence of psychosis and neurological involvement overall [47,48]. In the Hopkins cohort no sex differences have been found about seizures or psychosis, but cognitive impairment has been more frequently observed among African American men [49].

Whether different NP phenotypes emerge within different ethnicities is still a matter of debate [50]. Hispanics, African descendants and Asians were found to develop NP abnormalities more frequently than Caucasians [50e56]; Asian lupus patients also have worse disease compared with Caucasians and they have a higher rate of NP disease than Caucasians [54,55] and comparable to that reported in Afro-Caribbean patients [57].

According to the LUMINA study [58] older age is associated with more frequent appearance of cognitive dysfunction and

Please cite this article in press as: M. Govoni, et al., The diagnosis and clinical management of the neuropsychiatric manifestations of lupus,

Journal of Autoimmunity (2016), Downloaded from at GOT-CILEA-Anna - Ferrara Hospital Library July 18, 2016.

For personal use only. No other uses without permission. Copyright ?2016. Elsevier Inc. All rights reserved.

4

M. Govoni et al. / Journal of Autoimmunity xxx (2016) 1e32

Table 2 Frequency of neuropsychiatric syndromes as reported in cohort studies using the ACR nomenclature and case definitions for NPSLE [20,23,31e36].

%

NP syndromes

Common Frequent Uncommon Rare

>10e20 5e10 3e5 7 provided a PPV of 86.3% (95% CI 76.2e93.2%), and a score 6 months before SLE onset)

0

Concomitant (within 6 months of SLE onset)

3

After (>6 months after SLE onset)

2

Item 2. Minor or not specific NP events as defined by Ainiala et al. [10]

Present (i.e. minor or common NP events as proposed by Ainiala et al. [10])a

0

Absent (i.e. NP events other than those proposed by Ainiala et al. [10])a

3

Item 3b. Confounding factors or not SLE-related associations as defined by the ACR glossary [8]

None or not applicable

2

Present (one confounding factor)

1

Present (more than one confounding factor)

0

Item 4b. Additional (or favouring) factors

None or not applicable

0

Present (one additional or favouring factor)

1

Present (more than one additional or favouring factor)

2

a List of NP pictures deemed as minor or common known to occur frequently in normal healthy population controls:

headaches, anxiety, mild depression (mood disorders failing to meet the criteria for major depressive-like episodes), mild cognitive impairment (deficit in fewer than three of the eight specified cognitive domains) and polyneuropathy without electrophysiological confirmation.

b A list of confounding and favouring factors is available in supplementary documents.

above has introduced two relevant new aspects: firstly it provides a scoring system and secondly it has included an item that takes into account also factors favouring the attribution.

Overall, this model has allowed a confident correct attribution of NP events deemed as SLE related in about 33% of cases, a percentage quite similar to that found in the SLICC cohort [1].

It is important to acknowledge that all the attribution models developed to date, at the best of their performances, leave an unsolved, and probably unsolvable, grey area of uncertainty, which however fits well with the objectively difficult clinical approach to bedside.

5. Risk factors for NPSLE

In the general management of patients with SLE and in order to warrant the best clinical monitoring strategy, it appears to be of great importance to identify those risk factors, which have been demonstrated to be related or predictive of a given manifestation of the disease. Similarly, to check a clinical and serological risk profile related to NP involvement in SLE, it seems to be of utmost importance in daily clinical practice to allow an appropriate surveillance, a prompt recognition and even a secondary prevention. To date, at least three major SLE-related risk factors have been reported and acknowledged by 2010 EULAR recommendations [9] as consistently linked to NPSLE; moreover these risk factors have been confirmed by some recent studies: a) generalized SLE disease activity and cumulative damage have been shown to be associated with an increased risk of seizures and severe cognitive dysfunction [69e72]; indeed, when patients with NP events attributed to SLE (particularly CNS diffuse NP events) were separated from those with NP event not SLE-related, SLE disease activity scores showed to be higher in the former [73]; b) previous or concurrent major NPSLE events, particularly stroke and seizures, showed to predict similar, future NP events [74e76]; in the large SLICC inception cohort, a greater risk of mood disorder was associated with concurrent neuropsychiatric events (p 0.01), while a lower risk was associated with Asian race/ethnicity (p ? 0.01) and treatment with immunosuppressive drugs (p ? 0.003) [77]. In another longitudinal study, patients who had suffered from antecedent NP involvement had significantly worse and persistently impaired memory and learning deficits compared to non-NPSLE patients over the 12month re-assessment period [78]; c) anti-phospholipid antibodies

(aCL, anti-b2GP1, LA) were associated with CVA [70,75,79,80], sei-

zures [23,69,74], myelopathy [81], chorea [23], movement disorders [23] and moderate to severe cognitive dysfunction [23,71,82]; in a multicentric italian cross-sectional study [61] of 959 SLE patients, aPL antibodies and/or antiphospholipid syndrome (APS) emerged as the strongest risk factor for primary NPSLE, particularly focal neuropsychiatric events. In the SLICC inception cohort, including more than 1000 SLE patients, assessed prospectively for up to 10 years, the presence of LA at baseline was associated with subsequent intracranial thrombosis; in this study, anti-ribosomal P antibodies were also identified as a risk factor for SLE-related psychosis, but other autoantibodies did not predict NP events [83].

Although headache is probably the most frequent NP manifestation in SLE, overall, it does not appear associated with any of the above mentioned factors such as global disease activity or specific autoantibodies [66].

Other SLE-related factors identified as linked to NP involvement were investigated and proposed by others, such as low levels of C3 and C4, presence of anti-Ro/SSA antibodies, vasculitis, nephritis, anti-dsDNA and anti-Sm antibodies, but with conflicting results [84e88].

Non-SLE-related factors such as increasing age, hypertension and other traditional atherosclerotic risk factors have been associated with cognitive dysfunction, depression and cerebrovascular disease [66,89e91]. From a practical point of view, a tight control of modifiable risk factors, a more intensive follow-up and a better awareness about the identification of a "risk profile" could optimize the management of NPSLE and translate into more targeted future preventive strategies for patients classified at higher risk for the development of NP events. However, to date, only a little preliminary evidence from prospective and retrospective cohort studies support a role for antiplatelet agents in the primary prevention of aPL associated thrombotic and cerebrovascular disease events [92e94].

6. Diagnosis

In the absence of specific and reliable biomarkers, diagnosing NPSLE is a difficult task, mainly based on expert opinion (still assumed as the current "gold standard") and firstly, a careful process of exclusion of causes not directly related to SLE is mandatory.

The 1999 ACR case definition of NPSLE clearly underscores this

Please cite this article in press as: M. Govoni, et al., The diagnosis and clinical management of the neuropsychiatric manifestations of lupus,

Journal of Autoimmunity (2016), Downloaded from at GOT-CILEA-Anna - Ferrara Hospital Library July 18, 2016.

For personal use only. No other uses without permission. Copyright ?2016. Elsevier Inc. All rights reserved.

6

M. Govoni et al. / Journal of Autoimmunity xxx (2016) 1e32

point: "Neuropsychiatric lupus erythematosus includes the neurologic syndromes of the central, peripheral and autonomic nervous system and the psychiatric syndromes observed in patients with SLE in which other causes have been excluded" [8].

A diagnostic approach to NPSLE was suggested by ACR classification - where indications for the ascertainment of each NP syndrome were provided [8] - and subsequently systematically translated into evidence and eminence (expert)-based recommendations by the task force of the EULAR standing committee for clinical affairs, in 2010 [9].

As a general rule, when dealing with new or unexplained NP symptoms or signs suggestive of neuropsychiatric disease occurring in SLE patients, the initial diagnostic work-up should be similar to that in non-SLE patients presenting with the same manifestations.

To assess general SLE activity may contribute to the attribution of NP events to the underlying SLE, especially when dealing with diffuse manifestations. A careful evaluation of cardiovascular and other risk factors is also advised [95e97].

Depending on the type of neuropsychiatric manifestation, the general assessment should include: cell blood count and biochemistry tests to check metabolic and/or electrolyte disturbances, urinalysis (to assess for extra-CNS disease activity), tests for thyroid function, vitamin deficiencies, etc., serological tests such as C3, C4, anti-dsDNA and anti-phospholipid antibodies, which are those more strictly associated with NPSLE, especially NP focal manifestations. To date it is shared opinion that other autoantibodies (e.g. anti-ribosomal P, anti-NR2) have limited diagnostic utility and not easily available everywhere; work-up for infectious diseases and careful history about use of medications, illicit drugs and substances, also considering drug to drug interactions, is advised.

Special tests should include: lumbar puncture & CSF analysis. These procedures are useful primarily to rule out CNS infection (Gram stain, culture, PCR); to evaluate e with a low grade of specificity - mild abnormalities ([ WBCs, protein, Y gluc) detectable in 30e40% of active NPSLE or to rule-in inflammatory NPSLE; IgG index is increased in up to 75% of diffuse NPSLE but cannot allow differential diagnosis from multiple sclerosis (MS). Some promising biomarkers (i.e. IL-6, anti-NR2) for inflammatory NPSLE are not yet available nor entered in clinical practice.

Special instrumental work up should include EEG, primarily useful for the diagnosis of seizure disorder yielding abnormalities in 70e80% during the active phase and epileptiform activity, which can be predictive for seizure recurrences with a PPV and NPV of 73% and 79%, respectively. In the context of other NP syndromes, EEG studies show lower sensitivity-specificity (50e60%). Neuropsychological tests (to evaluate cognitive function), nerve conduction studies & electromyography (to assess peripheral neurological disorders) and psychiatric assessment should be performed when indicated on the basis of clinical presentation. About neuroimaging, MRI has to be considered the reference technique to assess brain & spinal cord structure and function and to rule-out CNS infection, trauma, tumor, etc. The recommended MRI protocol (brain and spinal cord) has been suggested by EULAR recommendations (see below) [9].

If MRI is negative or MRI lesions do not correlate with neuropsychiatric manifestations, advanced neuroimaging techniques may be considered, based on availability and local expertise. Since no single imaging test covers all the different mechanisms leading to brain injury, multimodality imaging should be preferred [98]. As a general rule it could be useful to couple a structural (i.e MRI) with a functional neuroimaging study such as SPECT. In combination with MRI, this approach has demonstrated to show a prevalent negative predictive value in ruling out NPSLE when both techniques

are negative [99]. Other modalities (MRS, PET, PWI, etc.) are usually less available and less used in every day clinical practice.

To date a multidisciplinary expert consensus, involving rheumatologists, neurologists, psychiatrists and neuroradiologists, remains the best strategy to reach a confident diagnosis of NPSLE, keeping in mind that in some cases this goal can be at its best only presumptive; to reduce the risk of misdiagnosis a tight follow up with careful periodic re-examination is advised [95].

7. NPSLE pathogenesis

Nowadays no single pathogenic pathway is likely to account for the variety of the neuropsychiatric syndromes that can occur in SLE. Complex interrelated mechanisms, including genetic factors, vasculopathy, vascular occlusion, neuroendocrine-immune imbalance, tissue and neuronal damage mediated by autoantibodies, inflammatory mediators, bloodebrain barrier (BBB) dysfunction and direct neuronal cell death can be all involved [16,19,100e104].

7.1. Genetics

Although most investigations were underpowered to detect close genotypeephenotype relationships, some evidence suggests an increased abundance of genetic susceptibility factors, i.e. single nucleotide polymorphisms, in a number of candidate genes such as HLADRB1, IRF5, STAT4, BLK, TNFAIP3, TNIP1, FCGR2B, and TNFSF13 in patients with NPSLE compared with those without neurological involvement [105,106].

Mutations in TREX1, a gene related to the Aicardi-Goutieres syndrome [107], which encodes three-prime repair exonuclease 1 (also known as DNase III), have been identified in NPSLE patients. Polymorphisms in this gene are associated with manifestations of CNS involvement, such as seizures. Interestingly, TREX1-deficient mice develop a lethal autoimmune syndrome and increased levels of type 1 interferon, which has been deemed relevant to SLE pathogenesis [108,109]. The HLA-DRB1-04 genotype and STAT4 rs10181656 have been associated with ischemic CVD, independently from the effects of traditional CV risk factors and aPL antibody status [110,111].

7.2. Vasculopathy

Microvasculopathy, initially attributed to deposition of immune complexes, is now considered to arise from in situ complement activation. Autopsy studies have suggested that vasculopathy instead of true vasculitis - is involved in CNS damage. Small-vessel non-inflammatory vasculopathy, micro-vessel occlusion, multifocal microinfarcts, intracranial embolism, microhaemorrhages and cortical atrophy were the most common pathological findings [112e114].

7.3. Inflammatory mediators

Several cytokines and chemokines appear to be upregulated during active disease and downregulated after treatment in NPSLE [115]. In NPSLE a variety of pro-inflammatory cytokines and chemokines, known to promote intrathecal antibody production, recruit immune cells and modulate neurotransmitter release, have been found elevated in the serum and CSF [116]. Some of them (i.e. IL-10 and IL-6), were shown to correlate with disease activity assessed by anti-dsDNA antibody levels and SLEDAI score [117e121]. To better define the role of these cytokines in the pathogenesis of NPSLE and their possible utility as biomarker or therapeutic target further investigation is needed.

Two independent studies reported significantly higher levels of

Please cite this article in press as: M. Govoni, et al., The diagnosis and clinical management of the neuropsychiatric manifestations of lupus,

Journal of Autoimmunity (2016), Downloaded from at GOT-CILEA-Anna - Ferrara Hospital Library July 18, 2016.

For personal use only. No other uses without permission. Copyright ?2016. Elsevier Inc. All rights reserved.

M. Govoni et al. / Journal of Autoimmunity xxx (2016) 1e32

7

matrix metalloproteinase 9 (MMP-9) in serum and CSF of patients with NPSLE than in patients without neuropsychiatric symptoms [122,123]. MMP-9 levels were found to be correlated with both the volume of MRI-detected brain lesions and intrathecal levels of IL-6 and IL-8. Similarly, intrathecal levels of plasminogen activator inhibitor 1 (PAI-1) were substantially increased in patients with NPSLE compared with SLE patients who did not exhibit overt neuropsychiatric involvement. Again, levels of PAI-1 correlated with CSF, IL-6 and IL-8 and were associated with the presence of increased glial fibrillary acidic protein (GFAP) in CSF, a marker of neuronal damage [124,125].

Overall, these data suggest that, in NPSLE, the production of IL-6, IL-8, IL-10, monocyte chemoattractant protein 1 (MCP-1) and granulocyte colony-stimulating factor (G-CSF) inside the CNS could be relevant in the pathogenesis of some diffuse central manifestation of NPSLE (acute confusional state, anxiety disorder, mood disorder and psychosis) [126].

Studies in patients with SLE and lupus-prone mice have pointed out a possible role also for TNF-related weak inducer of apoptosis (TWEAK) and Fn14, a TNF receptor superfamily member 12A, in the pathogenesis of NPSLE because they are able to mediate the activation of cell proliferation, angiogenesis, inflammation and apoptosis [127,128].

In patients with SLE, CSF levels of B-cell-activating and survival factors were shown to have increased compared with healthy individuals. The B-Cell Activating Factor (BAFF) and A ProliferationInducing Ligand (APRIL) were found to have increased more than 200-fold and 20 fold, respectively. However, CSF levels of BAFF and APRIL did not correlate with serum levels of these cytokines. Again, APRIL levels, but not BAFF, had increased in the CSF of patients with NPSLE compared with SLE patients without CNS involvement. Further studies are needed in order to evaluate therapeutic strategies targeting BAFF, APRIL or their receptor TACI in NPSLE [129,130].

In NPSLE, nitric oxide (NO) pathway has been reported to be active as stress factor that might directly affect neural tissue. When produced in large quantities, NO provokes neuronal cell death through various mechanisms including direct nitrosylation of NMethyl-D-aspartate (NMDA) receptors that are relevant in neurotransmission [131]. Patients had high numbers of peripheral blood cells expressing TNF mRNA, which closely correlated with increased levels of NO metabolites in the CSF, and with disease severity. NO pathway can be modulated by inflammatory cytokines and upregulated by complement components, immune complexes and aPL antibodies [131,104].

7.4. Autoantibodies (Abs)

A number of autoantibodies have been implicated in NPSLE manifestations. Those targeting phospholipids (aPL), ribosomal P peptides (Rib-P), endothelial cell, glutamate receptor, microtubuleassociated protein 2, and MMP-9 are the most studied [63,117,132,133].

- Antiphospholipid antibodies (aPL). Lupus anticoagulant (LA),

anti-cardiolipin (aCL) and anti-b2GP1 antibodies, are the most

widely investigated autoantibodies. They have been largely recognized as one of the most relevant risk factor for NPSLE and as a potential direct contributor to the development of thrombosis and other NPSLE manifestations such as seizures, stroke, chorea, movement disorders, cognitive dysfunction and myelopathy [23,68,69,74,81,82]. The anti-phospholipid effects cannot be simply explained by their pro-thrombotic activity. A direct effect on neuronal cells was suggested by studies in vitro that have shown aPL binding with CNS cells and in vivo

induction of cognitive defects in mice due to intrathecal passive transfer of class G immunoglobulins (IgG) from patients aPL positive [134]. - Anti-ribosomal P protein antibodies (anti-Rib-P) are quite specific for SLE occurring in up to 46% of patients, and target epitopes located in the C-terminal end of three highly conserved phosphorylated proteins, P0, P1 and P2, which are present in the 60S subunit of ribosomes. While in most of retrospective crosssectional studies (see Table 4) an association between elevated serum or CSF titres of anti-ribosomal P protein antibodies and NPSLE was found to be controversial, in longitudinal and prospective studies [135e138] their association with lupus psychosis was observed. In a large inception cohort of 1047 SLE patients with a mean follow-up of 3.6 years, aimed to determine which Abs at enrolment predicted subsequent NP events, antiRib-P antibodies demonstrated to be predictive of psychosis with a Cox proportional hazards regression of 3.92 [63].

In healthy mice, anti-Rib-P were shown to recognize neurons in the hippocampus, cingulate and primary olfactory piriform cortex, and to induce a long-term depressive-like behaviour when administered into cerebral ventricula [155]. In an independent study, the neuropathogenic potential of anti-ribosomal P protein antibodies was also demonstrated [156]. Anti-Rib-P antibodies from patients with SLE, induced a rapid and sustained increase in calcium influx and apoptosis in rat neurons expressing a cell-surface P-antigen protein named p331. The death of these neurons in specific brain regions (i.e. hippocampus) involved in memory and emotional behaviour in rats, could potentially account for a broad range of deficits observed in patients with NPSLE [156,157]. To summarize, while in humans direct demonstration of a pathogenetic role of these antibodies is still lacking, experimental data in mice and longitudinal association-studies support their role in diffuse NPSLE. Their search could be useful in the attribution process for some NP events, such as psychosis [152]. - Anti-glutamate receptor antibodies (anti-NMDA or anti-

NR2) occur in 30e40% of patients with SLE and have been described as a subset of anti-double-stranded-DNA (dsDNA) antibodies cross-reacting with NMDA receptors [158], specifically with the NMDA receptor subunit 2 (NR2; also known as

Table 4 Association between anti-P-Abs and NPSLE.

Authors, year, references

Cross-sectional studies Golombek SJ et al., 1987 [139] Van Dam A et al., 1991 [140] Schneebaum AB et al., 1991 [141] Teh LS et al., 1992 [142] Teh LS et al., 1993 [143] Nojima Y et al., 1992 [144] Yoshio T et al., 1995 [145] Arnett FC et al., 1996 [146] Press J et al., 1996 [147] Isshi K et al., 1996 [148] Georgescu L et al., 1997 [149] Tzioufas AG et al., 2000 [150] Gerli R et al., 2002 [151] Karassa FB et al., 2006 [152] Abdel-Nasser AM et al., 2008 [153] Haddouk S et al., 2009 [154] Prospective studies Bonfa E et al., 1987 [135] West SG et al., 1995 [136] Watanabe T et al., 1996 [137] Briani C et al., 2009 [138] Hanly JG et al., 2008 [63]

Association yes/no

no no yes no no yes yes yes yes yes yes yes no no yes no

yes yes yes yes yes

Please cite this article in press as: M. Govoni, et al., The diagnosis and clinical management of the neuropsychiatric manifestations of lupus,

Journal of Autoimmunity (2016), Downloaded from at GOT-CILEA-Anna - Ferrara Hospital Library July 18, 2016.

For personal use only. No other uses without permission. Copyright ?2016. Elsevier Inc. All rights reserved.

8

M. Govoni et al. / Journal of Autoimmunity xxx (2016) 1e32

GluN2) exerting an excitotoxic effect. NMDA receptors are widely distributed throughout the

brain, localized within glutamatergic synapses, with a particularly high density observed in the amygdala and hippocampus, two cerebral sub-regions implicated in cognitive functions, emotional processes and memory [159e162]. Activation of the NMDAR is critical in learning and memory, but prolonged stimulation can result in apoptotic death of the neuron.

The potential pathogenic role of anti-NMDA IgG antibodies detected in CSF has been demonstrated both in vitro and in vivo [163,164]. Clinical studies have shown a correlation between CSF anti-NMDA antibodies with diffuse CNS manifestations even if robust evidence of an association between anti-NR2/dsDNA antibodies and clinical pictures of NPSLE has not yet been demonstrated [165e167].

It has been suggested that anti-NR2/dsDNA antibodies may also distinguish SLE patients with central, diffuse CNS manifestations from patients with PNS involvement or without neuropsychiatric manifestations, at all [168e170]. Furthermore, a synergism between anti-NMDA and aPL antibodies has been hypothesized in inducing tissue damage and cognitive dysfunction [171]. - Anti-endothelial cell antibodies (AECA). Their frequency in SLE patients ranges from 17% to 75%. An association between serum AECA with psychosis and depression has been reported. AECAs recognize constitutively expressed or other cryptic antigens that are cytokine-induced, as well as adhesion molecules [172].

Other autoantibodies have been linked to NPSLE but need further supportive studies: antibodies targeting MAP-2, a cellular protein essential to cytoskeletal integrity found, almost exclusively, in neurons [173]; according to some studies anti-U1RNP antibodies might represent a clinically important CSF biomarker being even more specific for NPSLE than anti-ribosomal P protein or antiNR2/dsDNA antibodies [174e176]; anti-PARP-1 antibodies recognize poly (ADP-ribose) polymerase 1 (PARP-1) and are hypothesized to be involved in single-stranded DNA repair. These antibodies are significantly less frequently observed in patients with NPSLE than in patients with SLE without neurological or renal features [175,177,178].

Clustering SLE patient by auto-antibody profile may be useful in order to identify clinical subsets and to monitor disease progression [179,116]. (see Table 5: Association between Abs and specific neuropsychiatric syndromes).

Compared with SLE patients, NPSLE patients are more likely to have elevated serum levels of LA, APL, aCL [23,27,37,63,70,84,84,89,136,136,172,179e185], anti-Rib-P Abs [63,136,141e144,146,149,150,153,172,179,180,184,186e189] and increased CSF levels of anti-neuronal Abs [136,184,190e194]. Table 5 summarizes the common association between NPSLE clinical pictures and single Abs positivity or combined Abs positivity.

7.5. Bloodebrain barrier (BBB) dysfunction and cerebrospinal fluid autoantibodies

It is still unclear whether local intrathecal antibody production or, most likely, their migration inside the CNS from blood circulation through increased permeability of the BBB, or even both mechanisms, take place in NPSLE. Evidence for BBB disruption or dysfunction in neurodegenerative diseases, as well as in NPSLE is accumulating but the mechanism by which self-reactive autoantibodies pass through the BBB and affect neurological functions remains poorly understood [195e197]. CSF analysis from NPSLE patients have repeatedly demonstrated increased levels of immunoglobulins, pro-inflammatory cytokines, and albumin, indicative of increased BBB permeability [174,198,199].

An elevated CSF IgG index (CSF IgG: serum IgG ratio) has been more likely attributed to intrathecal antibody synthesis. In BXSB and MRLlpr/lpr lupus-prone mice, an increased IgG index and elevated albumin concentrations were reported in CSF and it was shown to correlate with neurodegeneration in periventricular areas and disease activity [200e203].

In NPSLE patients, CSF autoantibodies levels have been shown to more closely correlate with psychiatric manifestations and brain injury than serum levels do [165,166,179,187,204,205]. Endothelial cells in the brain express NMDA receptors, and binding of glutamate to these receptors has been shown to result in loss of BBB integrity. Antibodies that recognize NR2 might promote their own transport into the brain by acting as agonists or co-agonists for NMDA receptors. Anti-NR2/dsDNA antibodies from patients with SLE have also been shown to bind human umbilical vein endothelial cells and to upregulate endothelial expression of ICAM-1 and VCAM-1, as well as the production of IL-6 and IL-8. Similarly, activation of BBB endothelial cells by anti-NR2/dsDNA antibodies might cause inflammation, BBB disruption, and entry of antibodies into the CSF in patients with SLE [206,167].

Evidence exists that the complement system (in particular, C5a/ C5aR) has a key role in the disruption of BBB integrity through different cascades of events leading to inflammation, increased generation of inducible nitric oxide synthase (iNOS), reactive oxygen species, and actin reorganization. These results suggest that inhibition of C5aR could be a future potential treatment strategy for NPSLE [207].

Activation of endothelial cell can be induced by proinflammatory cytokines or autoantibodies. The upregulation on their surface of adhesion proteins, such as ICAM-1 and VCAM-1, can facilitate leukocyte and immunocompetent cell migration inside the CNS. Serum levels of soluble ICAM-1 have been shown to correlate with disease activity and normalize with remission, furtherly supporting the hypothesis that endothelial cell activation and consequent compromise of BBB integrity could be an essential step in NPSLE pathogenesis [208,209].

Whereas stimulating, it should be kept in mind that many of the above reported observations are only supportive for some clinical associations whilst a close mechanistic involvement of particular

Table 5 Association between Abs and specific neuropsychiatric syndromes.

Autoantibody

Neuropsychiatric syndromes

aCL (serum) LA (serum) aPL (serum) Anti-ribosomal P (serum) Anti-ribosomal P (CSF) Anti-NMDA (serum)

Cerebrovascular disease, seizures, headache, movement disorder, cognitive dysfunction, cranial neuropathy Cerebrovascular disease, seizures, cognitive dysfunction Cerebrovascular disease, cognitive dysfunction Psychosis, Seizures, acute confusional status, mood disorder, movement disorder Seizures Cognitive dysfunction

Please cite this article in press as: M. Govoni, et al., The diagnosis and clinical management of the neuropsychiatric manifestations of lupus,

Journal of Autoimmunity (2016), Downloaded from at GOT-CILEA-Anna - Ferrara Hospital Library July 18, 2016.

For personal use only. No other uses without permission. Copyright ?2016. Elsevier Inc. All rights reserved.

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download