Lupus

http://lup.sagepub.com/

Neuropsychiatric manifestations in pediatric systemic lupus erythematosus: a 20-year study

H-H Yu, J-H Lee, L-C Wang, Y-H Yang and B-L Chiang
Lupus 2006 15: 651
DOI: 10.1177/0961203306070990

The online version of this article can be found at:

http://lup.sagepub.com/content/15/10/651

Published by:

http://www.sagepublications.com

Additional services and information for Lupus can be found at:

Email Alerts: http://lup.sagepub.com/cgi/alerts

Subscriptions: http://lup.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations: http://lup.sagepub.com/content/15/10/651.refs.html

>> Version of Record - Oct 11, 2006

What is This?

Downloaded from lup.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on April 9, 2014

 Lupus (2006) 15, 651–657
www.lupus-journal.com

PAPER

Neuropsychiatric manifestations in pediatric systemic lupus

erythematosus: a 20-year study
H-H Yu, J-H Lee, L-C Wang, Y-H Yang and B-L Chiang*
Department of Pediatrics, National Taiwan University Hospital, Taipei, Taiwan, Republic of China

The objective of this study was to investigate the manifestations, treatment and outcome of
neuropsychiatric (NP) involvement in pediatric systemic lupus erythematosus (SLE) patients. The
charts of 185 pediatric patients with SLE diagnosed between 1985 and 2005 in a tertiary referral
hospital were retrospectively reviewed. NPSLE were defined using the American College of
Rheumatology NPSLE case definitions. NPSLE developed in 34.6% (64/185) of the patients. The
mean onset age was 15.2 years. Fourteen patients (21.9%) had NP manifestations on initial diagnosis
of SLE. The median duration from the onset of SLE to NP manifestation was 11 months. The most
frequent NP manifestations were seizure disorder (84.4%), ischemic stroke (28.1%) and psychosis
(21.9%). However, the prevalence of manifestations of NPSLE might be underestimated by the
retrospective design of our study. Higher mean C3/C4 levels, less percentage of anti-dsDNA
antibodies elevation and higher percentage of elevated anticardiolipin antibodies were observed in
NPSLE events than in non-NPSLE events (P
0.05). The mortality rate of NPSLE patients
decreased from 52.2% in 1985–1994 cohort to 27.8% in 1995–2005 cohort. In the past 10 years, the
leading cause of death in NPSLE patients was NPSLE itself. NPSLE is common in pediatric SLE
patients. It has diverse manifestations and a high mortality. Lupus (2006) 15, 651–657.

Key words: systemic lupus erythematosus (SLE); neuropsychiatric manifestations

Introduction This review was the largest retrospective survey of

pediatric NPSLE patients. This study provides physi-
Systemic lupus erythematosus (SLE) is an autoim- cians some information about diagnosis and manage-
mune disease with multiple organ involvement. ment of pediatric NPSLE patients.
NPSLE remains a challenging disease in children. In
1999, the American College of Rheumatology (ACR)
revised a standard nomenclature and a set of case
definitions for 19 neuropsychiatric syndromes of
Materials and methods
NPSLE.1 NPSLE has been reported to affect 22–43%
We retrospectively reviewed the charts of pediatric
of pediatric SLE patients in retrospective studies2–5
SLE patients who attended the Pediatric
and 95% in Sibbitt’s prospective study.6 The pathoge-
Rheumatology Clinic of the National Taiwan
nesis has been related to vasculitis and several auto-
University Hospital from 1985 to 2005 and who satis-
antibodies, such as anti-ribosomal P antibodies,7
fied the American College of Rheumatology (ACR)
anti-neuronal antibodies8 and anti-phospholipid anti-
1997 revised criteria for SLE.14 All patients were
bodies.9 NPSLE was described primarily in adults
ethnic Chinese.
and only a few studies with a limited number of cases
The American College of Rheumatology (ACR)
have focused on children.2–6,10,12,13 We reviewed the
nomenclature and standard definitions for NPSLE1
incidence, clinical features, diagnostic evaluation,
was used to verify the clinical manifestations in our
treatment and outcome of NPSLE in pediatric patients.
patients. These included a diagnostic guideline for 19
NP symptoms. Patients were excluded from the study
when their NP manifestations were secondary to other
*Correspondence: Dr Bor-Luen Chiang, Department of Pediatrics,
National Taiwan University Hospital, No. 7 Chung-Shan South Road,
causes, such as hypertensive encephalopathy, uremia,
Taipei, Taiwan, Republic of China. E-mail: gicmbor@ha.mc.ntu.edu.tw infection or other central nervous system disease not
Received 8 February 2006; accepted 19 June 2006 related to SLE.
© 2006 SAGE Publications 10.1177/0961203306070990

Downloaded from lup.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on April 9, 2014

 Neuropsychiatric manifestations in pediatric SLE
H-H Yu et al.
652
Demographic, clinical features, diagnostic evalua- Table 1 Clinical features of SLE patients
tion, treatment and outcome of SLE patients were Total SLE patients NPSLE patients
recorded. Diagnostic evaluation included complete
blood and differential counts, anti-nuclear antibody, No of patients 185 64
Female : male 158 : 27 55 : 9
anti-dsDNA antibody, serum C3 and C4 levels, anti- Age of SLE diagnosed (years) 13.2  3 (4.3–18) 13  3 (4.3–18)
extranuclear antibody, anti-phospholipid antibody, Disease duration (years) 6.8  5.6 7.6  6.3
anti-cardiolipin antibody, erythrocyte sedimentation NPSLE 64 (34.6%) 64 (100%)
Lupus nephritis 118 (63.8%) 39 (60.9%)
rate (ESR) and c-reactive protein (CRP). SLICC 1.3  1.8 2.6  2
Antiphospholipid antibodies (APA) was assayed using
IMUCLONE® aPL IgG ELISA kit. The cutoff point Mean  standard deviation, number (percentage).
for positivity is set at 15 U/mL. Borderline positive is Abbreviations: SLICC  Systemic Lupus International Collaborating
Clinics (SLICC) damage index.
5–15 U/mL. Anticardiolipin antibodies (ACA) were
assayed using AUTOZYME® ACL anticardiolipin IgG
and IgM sandwich immunoassays. The cutoff point for
positivity is set at 20 U/mL. Borderline positive is (60.9%) also developed lupus nephritis in the course of
13.5–20 U/mL. Cerebral spinal fluid (CSF) study, SLE. The mean age onset of SLE was 13.2 years
brain (or spinal cord) imaging, such as computed (range: 4.3–18 years) and the female to male ratio was
tomography (CT), magnetic resonance image (MRI), 5.9 : 1. The mean onset age of NPSLE was 15.2 years
single photon emission computed tomography (range: 5–28.7 years), while the median period
(SPECT) and electroencephalogram (EEG) were between onset of SLE and NP manifestations was 11
performed whenever necessary. months (range: 0–16.6 years).
Global disease activity was quantified by the SLE Fourteen patients (21.9%) had NP manifestations
Disease Activity Index (SLEDAI), while cumulative on initial diagnosis of SLE, which were classified as
organ damage was assessed by the ACR/Systemic eight seizure disorder, two confusion, two headache,
Lupus International Collaborating Clinics (SLICC) two polyneuropathy, one CVA, one depression, one
damage index. psychosis, one demyelinating syndrome and one
Lupus nephritis or nephropathy was defined by the myopathy. Twenty-one patients (32.8%) developed
presence of any of the following indicators: proteinuria NP symptoms within the first year of SLE, while 29
0.5g/day, cellular cast, glomerular infiltration (45.3%) only after one year of SLE diagnosis.

50%, abnormalities on renal biopsy, and end-stage Sixty-four patients had a total of 223 NP events,
renal disease treated by transplant or dialysis. with an average of 3.5 events per person (Table 2).
To compare difference of the treatment and mortal- Ninety-six percent of patients had central nervous
ity, all of the patients were divided into two cohorts: system manifestation compared to peripheral nervous
the 1985–1994 and the 1995–2005 cohorts based on system involvement (3.8%). The most frequent NP
the time point when patient developed NPSLE. manifestations were seizure disorder (84.4%, with
67.2% generalized seizure and 17.2% partial seizure),
ischemic stroke (28.1%), and psychosis (21.9%).
Statistical analysis Fifty-four patients had seizure in a total of 133 seizure
Comparisons were done using the 2 test for categori- events, with an average of 2.5 episodes per patient.
cal variables. For continuous data, Student’s t-test Recurrence of seizure was observed in 27/43 (62.8%)
was employed. When the data did not follow a normal patients with generalized seizure and 3/11 (27.3%)
distribution, the Mann-Whitney U-test was used. patients with partial seizure. Status epilepticus was
Statistical significance was defined as P
0.05 and all observed in four patients, two of whom died. Three
of the statistics were computed using the SPSS pro- required long-term anti-convulsant therapy over six
gram (version 12.0 for Windows). months. Thirty-four percent of seizure occurred con-
currently with cardiovascular accident (CVA) and
8.5% with psychosis.
Results Twenty-five patients (39%) had CVA, one
bilateral amaurosis fugax, 18 ischemic stroke, four ICH
The basic demographic data of total SLE patients were (including two subarachnoid hemorrhage (SAH), one
summarized in Table 1. The study population was pre- subdural hemorrhage (SDH), and one intracranial hem-
dominantly female (85.4%) with mean disease dura- orrhage (ICH) due to rupture of arterio-venous malfor-
tion of 6.8 years. Out of 185 SLE patients, 64 (34.6%) mation), and two dural sinus thrombosis. Five (20.8%)
developed NP manifestations and 118 (63.8%) had CVA occurred alone, while CVA combined with
lupus nephritis (Table 1). Thirty-nine NPSLE patients seizure occurred in 18 (75%) patients.
Lupus

Downloaded from lup.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on April 9, 2014

 Neuropsychiatric manifestations in pediatric SLE
H-H Yu et al.
653
Table 2 Neuropsychiatric manifestations in 64 of 185 SLE patients cord with presentation of progressive quadriplegia,
NP manifestation Patients, n (%) Events, n urinary retention and respiratory failure. After four
cycles of MPT and two CYC pulse therapy, one intra-
Acute confusional state 8 (12.5 %) 8 thecal therapy with methotrexate and hydrocortisone,
Acute inflammatory demyelinating 0 0
polyradiculopathy and six monthly intravenous immunoglobulins (IVIG)
Anxiety disorder 4 (6.3%) 4 (2 g/kg/day), movement in only one arm and
Aseptic meningitis 0 0 respiratory function improved after eight months of
Cerebrovascular disease 25 (39%) 25
Transient ischemic stroke 1 (1.6%) 1 follow-up.
Ischemic stroke 18 (28.1%) 18 We compared the laboratory features in neuropsy-
Intracranial hemorrhage 4 (6.3%) 4 chiatric events and non-NPSLE events, but patients
Sinus thrombosis 2 (3.1%) 2
Cognitive dysfunction 0 0 with NPSLE with lupus nephritis at the same time
Demyelinating syndrome 2 (3.1%) 3 were excluded (Table 3). Higher mean C3/C4 levels
Headache 7 (10.9%) 6 (C3 54.8  26.1 versus 45.6  21.6 mg/dL, C4
Mononeuropathy 2 (3.1%) 2
Mood disorder median 9.1, range 4–55.8 mg/dL versus median
Depression 8 (12.5%) 9 8 mg/dL, range 1.2–56 mg/dL), less percentage of
Movement disorder (chorea) 0 0 anti-dsDNA antibodies elevation (32.6% versus
Myasthenia Gravis 0 0
Myelopathy 3 (4.7%) 3 90.9%) and higher percentage of elevated anticardi-
Transverse myelopathy 2 (3.1%) 3 olipin antibodies (47.8% versus 19.4%) were observed
Longitudinal myelopathy 1 (1.6%) 1 in NPSLE events than in non-NPSLE events
Neuropathy, cranial 1 (1.6%) 1
Neuropathy, autonomic 1 (1.6%) 2 (P
0.05). NPSLE events with normal anti-DNA and
Plexopathy 0 0 C3 and C4 levels were seizures, psychosis, depression,
Polyneuropathy 3 (4.7%) 3 autonomic dysfunction, headache, longitudinal
Psychosis 14 (21.9%) 21
Seizure disorder 54 (84.4%) myelitis and CVA.
Generalized 43 (67.2%) 119 In initial investigation of NPSLE events, CSF study
Partial 11 (17.2%) 14 abnormality was noted in 16/30 (53.3%) of patients
Total 132 223
with 11/16 (68.8%) WBC count elevation and 16/16
(100%) protein elevation in CSF. Computed tomogra-
One 10-year old girl had encephalitis and transverse phy (CT), magnetic resonance imaging (MRI) and
myelitis involving the medulla, pons, and thoracic single photon emission computed tomography
spinal cord (T2-T6 and T10-T11 levels). After methyl- (SPECT) abnormality was observed in 88.6%, 92.5%
prednisolone pulse therapy (MPT) and cyclosporine and 85.7% of patients, respectively. The most common
treatment, most neurologic deficits resolved. But brain CT/MRI abnormalities were brain atrophy and
transverse myelitis (involving the spinal cord at T1-T4 CVA. Sixty percent of patients had abnormal
and T7-L1 levels) relapsed three months later. ophthalmologic examination, with vasculitis as the
Paraplegia with neurogenic bladder persisted after two most common finding (Table 4).
cycles of MPT and one cyclophosphamide (CYC) Cumulative treatment of NPSLE patients in the
pulse therapy. 1985–1994 cohort and 1995–2005 cohort was shown
Another five-year -old girl had longitudinal myelitis in Table 5. MPT was used for patients with severe NP
involving the cervical, thoracic and lumbar spinal manifestations. A SLEDAI score 30 correlated to

Table 3 Comparison of disease activity markers and antibodies of neuropsychiatric events and
non-NPSLE events (exclude patients with NPSLE and lupus nephritis at the same time)
NPSLE Non-NPSLE P

C3 decrease (
80 mg/dL) 45/53 (84.9%) 111/119 (93.3%) 0.081

C4 decrease (
16 mg/dL) 39/52 (75%) 98/118 (83.1%) 0.221
Anti-dsDNA elevation 15/46 (32.6%) 100/110 (90.9%) 0.000*
C3 (mg/dL) 54.8  26.1 45.6  21.6 0.021*
C4 (mg/dL) 9.1 (4–55.8) 8 (1.2–56) 0.043*
Anti-DNA Ab (IU/mL) 45.4 (0–675.9) 97 (0.02–789.0) 0.001*
ESR (mm/hour) 64.3  37.2 64.2  37.4 0.965
CRP (mg/dL) 0.6 (0.04–25.1) 0.4 (0–6.9) 0.011*
ANA  1 : 1280 18/45 (40%) 82/119 (68.9%) 0.001*
Antiphospholipid Ab 8/23 (34.8%) 7/28 (25%) 0.446
Anticardiolipin Ab 11/23 (47.8%) 6/31 (19.4%) 0.026*

Abbreviations: ANA  anti-nuclear antibodies; ESR  erythrocyte sedimentation rate; CRP  c-reactive protein.
*P
0.05.

Lupus

Downloaded from lup.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on April 9, 2014

 Neuropsychiatric manifestations in pediatric SLE
H-H Yu et al.
654
Table 4 Investigation in 223 neuropsychiatric events of SLE
Investigation Number* (%) Classification

CSF study abnormality 16/30 (53.3%) CSF WBC elevation 11, protein elevation 16
EEG abnormality 50/71 (70.4%) Epileptic spikes 13, slow waves 41
CT abnormality 39/44 (88.6%) Brain atrophy 19, ICH 4, SAH 2, sinus thrombosis 1,
ischemia 14, basal ganglion calcification 2
MRI abnormality 37/40 (92.5%) Brain atrophy 14, ICH 3, infarction 10, myelopathy 3,
brainstem encephalitis 1, demyelinating encephalitis 1,
acute necrotizing encephalopathy 1, multifocal lesion
(small vessel vasculitis) 2
SPECT abnormality 6/7 (85.7%) Decreased brain perfusion in 4 seizure and 2 psychosis episodes
Opthalmology abnormality 35/51 (60.8%) Retinopathy 25, papilledema 5, glaucoma 5

*Number of abnormal/total number of investigations.

Abbreviations: CSF  cerebral spinal fluid; EEG  electroencephalography; CT  computed tomography; MRI  magnetic reso-
nance image; SPECT  single photon emission computed tomograph;, ICH  intracranial hemorrhage; SAH  sub-arachnoid
hemorrhage.

the use of MPT significantly (odds ratio: 3.8, 95% Overall mortality rate of total 185 SLE patients was
confidence interval: 1.3–11.5, P  0.014). The use of 27% (50/185), while 45.3% (29/64) for NPSLE patients
CYC pulse therapy and IVIG did not correlate with and 17.4% (21/121) for patients without NPSLE. The
SLEDAI significantly (P  0.05). There was no differ- mortality rate of NPSLE patients of the 1985–1994
ence in using MPT, CYC pulse therapy and IVIG in cohort was 52.2%, which was less (27.8%) in the
the two patient cohorts. For maintenance therapy, all of 1995–2005 cohort (P  0.099) (Table 6). The mortality
the patients with NPSLE were given corticosteroid rate of total SLE patients without NPSLE significantly
treatment. The next common oral medications were decreased from 34.5% in the 1985–1994 cohort to 3%
azathioprine, non-steroidal anti-inflammatory drugs in the 1995–2005 cohort (P  0.000). In 1985–1994
(NSAID), and anti-malarial agents (hydroxychloro- cohort, the major causes of death of NPSLE patients
quine). Compared with the 1985–1994 cohort, a higher were infection (70.8%), renal failure (62.5%), NPSLE
percentage of patients in the 1995–2005 cohort were (29.2%) and pulmonary hemorrhage or acute respira-
given hydroxychloroquine, salfasalazine and tory distress syndrome (ARDS) (29.2%). The infection
cyclosporine (P
0.05). rate decreased from 70.8% in the 1985–1994 cohort to

Table 5 Accumulative treatment in 64 patients with neuropsychiatric manifestations

1985–1994 cohort 1995–2005 cohort
Treatment n  46 n  18 P-value

NSAID 21 (45.7%) 11 (61.1%) 0.266

Antimalarials 16 (34.8%) 14 (77.8%) 0.002*
Oral corticosteroids 46 (100%) 18 (100%) —
Azathioprine 30 (65.2%) 13 (72.2%) 0.592
Salfasalazine 2 (4.3%) 4 (22.2%) 0.027*
Cyclophosphamide 11 (23.9%) 3 (16.7%) 0.528
Cyclophosphamide pulse therapy1 24 (52.2%) 7 (38.9%) 0.339
(CYC pulse)
CYC pulse times 4.5 (1–16) 6 (1–12) 0.633
Methylprednisolone pulse 27 (58.7%) 13 (72.2%) 0.315
therapy2 (MPT)
MPT times 2 (1–7) 3 (1–8) 0.143
Cyclosporin 4 (8.7%) 7 (38.9%) 0.004*
IVIG3 10 (21.7%) 4 (22.2%) 0.966
IVIG times 1 (1–2) 2 (1–6) 0.091
Mycophenolate mofetil 1 (2.2%) 1 (5.6%) 0.485

Number (%) or median (range).

*Compare between two cohorts, P
0.05.
Abbreviations: NSAID: nonsteroid antiinflammatory drugs, IVIG: intravenous immunoglobulins.
1cyclophosphamide pulse therapy donates intravenous cyclophosphamide 0.5–1 g/m2.
2Methylprednisolone pulse therapy donates intravenous methylprednisolone 15–30 mg/kg/day for three days, maximum

1 g/day.
3IVIG donates intravenous immunoglobulins 2 g/kg/day.

Lupus

Downloaded from lup.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on April 9, 2014

 Neuropsychiatric manifestations in pediatric SLE
H-H Yu et al.
655
Table 6 Mortality in NPSLE and non-NPSLE patients and causes of deaths in patients with NPSLE
1985–1994 cohort 1995–2005 cohort P-value

Mortality of NPSLE patients 24/46 (52.2%) 5/18 (27.8%) 0.099

Mortality of SLE patients without NPSLE 19/55 (34.5%) 2/66 (3.0%) 0.000*
Cause of death of NPSLE patients
NPSLE 7/24 (29.2%) 3/5 (60%) 0.306
Infection 17/24 (70.8%) 1/5 (20%) 0.054
Pulmonary hemorrhage/ARDS 7/24 (29.2%) 1/5 (20%) 1.000
Renal failure 15/24 (62.5%) 1/5 (20%) 0.144
Unknown 3/24 (12.5%) 1/5 (20%) 0.444

Number (%).
*P
0.05, by Fisher’s exact test, two-tailed.

Figure 1 Survival curve of SLE patients with NP manifestations (NPSLE), lupus nephritis (LN), without NPSLE or LN (non-NP/LN)
(P  0.012 between NPSLE and LN, P  0.008 between NPSLE and non-NP/LN, P  0.20 between LN and non-NP/LN by the log-rank
test of Kaplan–Meier method).

20% in the 1995–2005 cohort (P  0.054). However, in (P  0.008 and 0.012). The 10-year survival rate for
the cohort of 1995–2005, the leading cause of death patients with NPSLE was 52.2%, and the 20-year
changed to NPSLE itself (60%). Causes of death due to survival rate was 31.8%. The mean survival was
NPSLE were five seizure disorders, including two status 12.1 years (95% confidence interval 9.7–14.5 years)
epilepticus, five CVA and one brainstem encephalitis. for NPSLE patients. Furthermore, the mortality rate of
The Kaplan–Meier survival curves of total SLE patients with NPSLE (11/25, 44%) and patients with
patients dividing into patients with NPSLE (NPSLE NPSLE and LN in the course of SLE (18/39, 46.2%)
group), patients with LN (LN group), and patients did not differ significantly (P  0.866).
without NPSLE or LN (non-NP/LN group) were To evaluate neurological outcome, the mean follow-
shown in Figure 1. There was significant difference by up duration was 7.6 years. Thirty-one patients recovered
the log-rank test of the Kaplan–Meier method between completely (48.4%), 10 patients had partial recovery
NPSLE and non-NP/LN, and NPSLE and LN group (15.6%) and others had unknown endpoints.
Lupus

Downloaded from lup.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on April 9, 2014

 Neuropsychiatric manifestations in pediatric SLE
H-H Yu et al.
656
Discussion Sanna et al.’s study of adult NPSLE patients,9 while
aPL antibodies are associated with ischemic CVA,
The prevalence of NP manifestations in SLE patients amaurosis fugax in pediatric NPSLE patients.17 Our
varies widely among different series. NPSLE have been data did not find any significant association between
reported to occur in 22–95% pediatric SLE patients and anti-phsopholipid antibodies and NPSLE or CVA,
15–81% of whom had NP manifestations at the onset of which might be due to small number of cases.
SLE.2–6 In our study, 34.6% patients had NP manifesta- EEG, CT or MRI abnormalities are also not specific
tions, while 21.8% occurred at SLE onset. In the previ- in NPSLE. SPECT is a sensitive but non-specific tool
ous report of pediatric NPSLE patients, the frequency in initial diagnosis of NPSLE in children.18–20 It is
of NP syndromes are as follows: headache helpful in patients with psychosis or diffuse brain
10–72%,2–6,20 cognitive disorder 55%,6 mood disorder involvement. Abnormal SPECT was detected in 85.7%
17–57%,2,6 seizure disorder 9–64%,2–6,20 CVA in our patients, which is similar to those of previous
0–30%,2–6 psychosis 4–50%,2,3,5,6,20 acute confusional reports (80–100% abnormality).18–20
state: 35%,6 anxiety 21%,6 peripheral nervous system Therapy with MPT and CYC pulse therapy appears
involvement 6.3–15%,3,6 chorea 7–20%,2,4–6 demyeli- effective for severe NPSLE.10,22–27 IVIG21 or plasma-
nating syndrome 4%6 and myelopathy 1–14%.2–6 The pheresis28 in combination of MPT, may be an alternative
most frequent NP manifestations of SLE were seizure to CYC pulse therapy. However, the management of
disorder (84.4%), ischemic stroke (28.1%) and psy- NPSLE has been entirely empirical and based on uncon-
chosis (21.9%) in our series, which is similar to a study trolled studies. None to 85% of patients received MPT
by Mok et al.11 of Chinese NPSLE patients who in acute NPSLE in the previous report.2–5 Considering
reported seizure disorder (28%), CVA (19%), acute the treatment, we basically followed McCune’s27 defini-
confusional state (14%) and psychosis (11%) being the tion of severe NPSLE: the presence of focal neurologi-
most common manifestations. cal deficits or seizures occurring in association with
Cognitive disorder is diagnosed by cognitive com- clinical evidence of disease progression, which was
plaints or objective cognitive dysfunction evaluated by compatible that MPT usage in acute NP events corre-
standardized neuropsychological tests, which were not lated with a SLEDAI score more than 30. We used
routinely performed in our patients. Other neuropsy- intrathecal administration of methotrexate and hydro-
chiatric manifestations, such as peripheral neuropathy cortisone in one patient with longitudinal myelitis with
or psychiatric disorder, also may not have been docu- limited therapeutic effect. Intrathecal administration of
mented without evaluation prospectively. Our study methotrexate has been reported as a possible treatment
likely underestimated the prevalence of NPSLE in for severe NPSLE unresponsive to conventional steroid
these groups of patients. The main limitation of our therapy by Valesini et al.29 and Dong et al.30
study is the retrospective review. Compared with the 1985–1994 cohort, a higher per-
Laboratory disease markers for SLE (anti-dsDNA, centage of patients in the 1995–2005 cohort used
C3 and C4 levels) were not consistently correlated hydroxychloroquine, salfasalazine, and cyclosporine as
with clinical flares of NPSLE. Elevation of anti- maintenance therapy (P
0.05). Although there was no
dsDNA antibodies was detected in 32.6% of NSPEL significant correlation between acute or maintenance
patients, which is still lower than the 75–89% in previous treatment and improved survival (data not shown), mor-
studies.3–5,10 Complement level decreased in 75 to tality decreased from 52.2% in 1985–1994 to 27.8% in
84.9% of our patients, which is consistent with those the 1995–2005 cohort, while infection, renal failure, and
of previous reports (36–75%).3–5,10 The results suggest pulmonary hemorrhage/ARDS decreased as causes of
that NPSLE can occur with a state of lower disease death. We believe that more aggressive treatment in
activity markers presented by C3, C4 and anti-dsDNA. severe NPSLE and lupus nephritis, more successful use
Absence of serological activity during active NPSLE of immunosuppresants with early corticosteroid taper-
has been proposed in previous studies by Shimojima ing, and better infection control in the past 10 years
et al.15 and Winfield et al.16 who demostrated that ele- contribute to improved survival.
vated anti-dsDNA antibodies levels and decreased Patients with NPSLE had higher mortality than LN
complement have been observed only in NPSLE and non-NP/LN patients in our study. We also showed
patients with extra-NP manifestations. that the mortality rate of patients combined with
CSF examination is extremely important to rule out NPSLE and LN were not significantly different to that
the possibility of infection. High protein with/without of patients with NPSLE only, which has not been
pleocytosis may be seen in 8–50% of NPSLE reported before in pediatric series. Our mortality rate of
patients,3–5 which is similar to our 53.3% CSF abnor- 45.3% is higher than the 4–35% mortality in previous
mality. Anti-phospholipid (aPL) antibodies are signifi- reports,2–5,13 which might be due to longer follow-up
cantly associated with CVA, headache, and seizures in duration in our series or ethnic variation. Neurologic
Lupus

Downloaded from lup.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on April 9, 2014

 Neuropsychiatric manifestations in pediatric SLE
H-H Yu et al.
657
recovery was excellent in the 84.4% of patients, while 12 Al-Mayouf SM, Bahabri S. Spinal cord involvement in pediatric sys-
temic lupus erythematosus: case report and literature review. Clin Exp
the 15.6% neurological sequelae in our patients is Rheumatol 1999; 17: 505–508 (review).
consistent with the previous report of 15–35.7%.3,5 13 Vyas S, Hidalgo G, Baqi N, Von Gizyki H, Singh A. Outcome in
In summary, NPSLE is common in children and has African-American children of neuropsychiatric lupus and lupus nephri-
tis. Pediatr Nephrol 2002; 17: 45–49.
diverse manifestations in clinical symptoms and sero- 14 Hochberg MC. Updating the American College of Rheumatology
logical findings. It is important to make an accurate revised criteria for the classification of systemic lupus erythematosus.
diagnosis of primary NPSLE and exclude secondary Arthritis Rheum 1997; 40: 1725–1726.
15 Shimojima Y, Matsuda M, Gono T, Ishii W, Ikeda SI. Relationship
causes, such as infection, hypertension, drugs, and between clinical factors and neuropsychiatric manifestations in systemic
metabolic derangement by CSF analysis, EEG, brain lupus erythematosus. Clin Rheumatol 2005; 24: 469–475.
imaging and serological tests of lupus activity and 16 Winfield JB, Brunner CM, Koffler D. Serologic studies in patients with
systemic lupus erythematosus and central nervous system dysfunction.
CRP. NPSLE can occur with a state of lower disease Arthritis Rheum 1978; 21: 289–294.
activity markers presented by C3, C4 and anti-dsDNA. 17 Campos LM, Kiss MH, D’Amico EA, Silva CA. Antiphospholipid
NPSLE remains the leading causes of mortality and antibodies and antiphospholipid syndrome in 57 children and
adolescents with systemic lupus erythematosus. Lupus 2003; 12:
morbidity in SLE. Treatment with methylprednisolone 820–826.
pulse therapy and cyclophosphamide pulse therapy in 18 Szer IS, Miller JH, Rawlings D, Shaham B, Bernstein B. Cerebral per-
severe NPSLE, together with better infection control fusion abnormalities in children with central nervous system manifesta-
tions of lupus detected by single photon emission computed
and treatment of concomitant medical condition, tomography. J Rheumatol 1993; 20: 2143–2148.
seems to improve neurologic outcome and survival. 19 Reiff A, Miller J, Shaham B, Bernstein B, Szer IS. Childhood central
nervous system lupus; longitudinal assessment using single photon
emission computed tomography. J Rheumatol 1997; 24: 2461–2465.
20 Russo R, Gilday D, Laxer RM, Eddy A, Silverman ED. Single photon
References emission computed tomography scanning in childhood systemic lupus
erythematosus. J Rheumatol 1998; 25: 576–582.
1 The American College of Rheumatology nomenclature and case defini- 21 Milstone AM, Meyers K, Elia J. Treatment of acute neuropsychiatric
tions for neuropsychiatric lupus syndromes. Arthritis Rheum 1999; 42: lupus with intravenous immunoglobulin (IVIG): a case report and
599–608 (review). review of the literature. Clin Rheumatol 2005; 24: 394–397.
2 Parikh S, Swaiman KF, Kim Y. Neurologic characteristics of childhood 22 Neuwelt CM, Lacks S, Kaye BR, Ellman JB, Borenstein DG. Role of
lupus erythematosus. Pediatr Neurol 1995; 13: 198–201. intravenous cyclophosphamide in the treatment of severe neuropsychi-
3 Yancey CL, Doughty RA, Athreya BH. Central nervous system involve- atric systemic lupus erythematosus. Am J Med 1995; 98: 32–41.
ment in childhood systemic lupus erythematosus. Arthritis Rheum 1981; 23 Takada K, Illei GG, Boumpas DT. Cyclophosphamide for the treatment
24: 1389–1395. of systemic lupus erythematosus. Lupus 2001; 10: 154–161.
4 Quintero-Del-Rio AI, Miller V. Neurologic symptoms in childrem with 24 Trevisani VF, Castro AA, Neves Neto JF, Atallah AN. Cyclophos-
systemic lupus erythematosus. J Child Neurol 2000; 12: 803–807. phamide versus methylprednisolone for the treatment of neuropsychi-
5 Olfat MO, Al-Mayouf SM, Muzaffer MA. Pattern of neuropsychiatric atric involvement in systemic lupus erythematosus. Cochrane Database
manifestations and outcome in juvenile systemic lupus erythematosus. Syst Rev 2000; 3: CD002265.
Clin Rheumatol 2004; 23: 395–399. 25 Sanna G, Bertolaccini ML, Mathieu A. Central nervous system lupus: a
6 Sibbitt WL Jr, Brandt JR, Johnson CR et al. The incidence and preva- clinical approach to therapy. Lupus 2003; 12: 935–942.
lence of neuropsychiatric syndromes in pediatric onset systemic lupus 26 Bodani M, Kopelman MD. A psychiatric perspective on the therapy
erythematosus. J Rheumatol 2002; 29: 1536–1542. of psychosis in systemic lupus erythematosus. Lupus 2003;
7 Eber T, Chapman J, Shoenfeld Y. Anti-ribosomal P-protein and its role 12: 947–949.
in psychiatric manifestations of systemic lupus erythematosus: myth or 27 McCune WJ, Golbus J, Zeldes W, Bohlke P, Dunne R, Fox DA. Clinical
reality? Lupus 2005; 14: 571–575. and immunologic effects of monthly administration of intravenous
8 Isshi K, Hirohata S. Differential roles of the anti-ribosomal P antibody cyclophosphamide in severe systemic lupus erythematosus. N Engl J
and antineuronal antibody in the pathogenesis of central nervous system Med 1988; 318: 1423–1431.
involvement in systemic lupus erythematosus. Arthritis Rheum 1998; 28 Neuwelt CM. The role of plasmapheresis in the treatment of severe
41: 1819–1827. central nervous system neuropsychiatric systemic lupus erythematosus.
9 Sanna G, Bertolaccini ML, Cuadrado MJ et al. Neuropsychiatric mani- Ther Apher Dial 2003; 7: 173–182.
festations in systemic lupus erythematosus: prevalence and association 29 Valesini G, Priori R, Francia A et al. Central nervous system involve-
with antiphospholipid antibodies. J Rheumatol 2003; 30: 985–992. ment in systemic lupus erythematosus: a new therapeutic approach with
10 Baca V, Lavalle C, Garcia R et al. Favorable response to intravenous intrathecal dexamethasone and methotrexate. Springer Semin
methylprednisolone and cyclophosphamide in children with severe neu- Immunopathol 1994; 16: 313–321.
ropsychiatric lupus. J Rheumatol 1999; 26: 432–439. 30 Dong Y, Zhang X, Tang F, Tian X, Zhao Y, Zhang F. Intrathecal injection
11 Mok CC, Lau CS, Wong RW. Neuropsychiatric manifestations and their with methotrexate plus dexamethasone in the treatment of central
clinical associations in southern Chinese patients with systemic lupus nervous system involvement in systemic lupus erythematosus. Clin Med
erythematosus. J Rheumatol 2001; 28: 766–771. J (Engl) 2001; 114: 764–766.

Lupus

Downloaded from lup.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on April 9, 2014