Childhood cerebral lupus in an Oriental population

Brain & Development 21 (1999) 229–235

Original article

Childhood cerebral lupus in an Oriental population Imam Haji Muhammad Ismail Hussain*, Wan Fei Loh, Ali Sofiah Paediatric Institute, Hospital Kuala Lumpur, Jalan Pahang, 50586 Kuala Lumpur, Malaysia Received 17 August 1998; received in revised form 30 December 1998; accepted 7 January 1999

Abstract In a cross-sectional study of 24 Oriental children with systemic lupus erythematosus (SLE)with a mean age of 11.25 years, 75% were found to have clinical and neurophysiological evidence of cerebral lupus. Seizures were the most common manifestation affecting 11 (61%) of the cases, followed by psychosis in five (27.7%), encephalopathy in five (27.7%) headaches in five (27.7%), personality changes in four (22.2%), stroke in three (16.6%), movement disorders in three (16.6%) and myelitis in one child (5.5%). Four children had cerebral lupus as the presenting manifestation of SLE. Twenty-one children had an electroencephalogram (EEG) of which 11 were normal. Abnormalities detected in the rest included focal sharps, slowing of background and electrodecremental changes. There was a poor correlation of EEG with the clinical presentation. Sixteen children with cerebral lupus had a computed tomogram (CT) of which three were normal. The commonest abnormality was cerebral atrophy with or without infarcts. Only four of the cases had lupus anticoagulant but compliment was reduced in 13. Sixteen of the cases also had renal involvement. Treatment was generally with steroids with only two patients receiving cyclophosphamide for cerebral relapse. Eight children (44%) made a full recovery. Learning disability was the most frequent sequelae affecting one-third of children seen at a 1-year follow up. Four (22%) had epilepsy, two (11%) had motor deficits and one child had optic atrophy. One child died of cerebral haemorrhage during a hypertensive crisis.  1999 Elsevier Science B.V. All rights reserved. Keywords: Cerebral lupus; Oriental children; Clinical features; Outcome

1. Introduction Systemic lupus erythematosus (SLE) is a multisystem disorder characterised by serum antibodies against a wide variety of nuclear, cytoplasmic and serum protein antigens. There is a female preponderance of 8:1 which is thought to be due to a synergistic effect of female hormones [1]. In addition there are racial differences in the incidences of SLE ranging from 31/100 000 among oriental females to 4.4/ 100 000 in white females, with black females falling in between at 19.8/100 000 [1]. The actual reason for this difference which is presumably genetic is unclear. In view of the protean manifestations, the American College of Rheumatology (ARA) has listed 11 diagnostic criteria and patients who have four of these are considered to have

* Corresponding author. Department of Paediatrics, Penang Hospital, 10450 Penang, Malaysia. Fax: +60-603-2948187; e-mail: [email protected]

0387-7604/99/$ - see front matter PII: S03 87-7604(99)000 12-1

SLE [2].These criteria were developed to distinguish SLE from the other causes of a positive anti-nuclear factor (ANF). In adult series [3], 10–75% of SLE patients were found to experience neurological complications with or without behavioural, cognitive and psychiatric disturbances leading to the term neuropsychiatric lupus. There have been few studies in children [4,5] and mostly in Western populations who have a lower incidence of SLE. We undertook a cross-sectional study to look at the spectrum of neurological complications in Oriental children with SLE.

2. Materials and methods All children with SLE referred to the Paediatric Institute of Hospital Kuala Lumpur in 1996 were selected for the study. The following aspects were studied: history of neuropsychiatric involvement, physical examination to document residual deficits, behavioural assessment, review of computed tomograms (CT)/magnetic reasonance imaging

 1999 Elsevier Science B.V. All rights reserved.

Age (d)

12 (13)

11 (4)

8 (14)

12 (36)

12 (18)

13 (20)

10 (27)

9 (22)

11 (24)

13 (25)

15 (62)

13 (12)

10 (8)

9 (24)

9 (35)

10 (26)

14 (47)

13 (26)

No.

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

F

F

F

M

F

F

F

F

F

F

F

F

F

F

F

F

F

F

Sex

Indian

Chinese

Malay

Chinese

Chinese

Malay

Chinese

Chinese

Malay

Chinese

Malay

Malay

Chinese

Malay

Dayak

Malay

Malay

Chinese

Race

Encephalopathy, Gen. fits Psychosis

Nil

Encephalopathy, Gen.fits* centro-temporal discharge Nil

Flaccid paraparesis Psychosis, incontinence Nil (mood swings) Nil

Encephlopathy

R. Hemichorea * two episodes Nil

R. Hemiplegia, R. focal fits Psychosis*

Gen. and focal fits

Encephlopathy, Gen. fits Headache

Coma * Gen. fits S.E. Cognitive/ mood swings R. Hemiplegia, Gen. fits Nil

CNS manifestations

discharges, focal slow

Bilateral epileptic

Normal

Normal

Diffuse delta, R.

Normal

Normal (ND in acute state) L. parietooccipital sharp waves, N. background Normal

Normal

Normal

Low voltage background

Bil.central sharp waves Normal

Normal

Normal

ND

Normal

Normal

R. Tibial dcv and amp. Both tibials dec. amp. Normal

Both tibials dcv

Normal

Normal

Normal

Normal

Normal

R. Tibial dcv and amplitude Both surals absent R. Tibial reduced amplitude R. Tibail dcv

Both tibails and surals dcv Normal

L. focal sharp waves Bilateral post. sharp waves R. focal sharp waves

Nerve conduction

Electroencephalogram

Summary of clinical, neurophysiological and radiological findings and outcome

Table 1

Multiple infarcts

ND

ND

Cerebral atrophy

Pred. Cyclo

Pred. Cyclo

Pred. Cyclo

Pred. Cyclo

Pred. Cyclo

Pred. Cyclo

Calcification MRI spine N ND ND

Pred.

Pred.

Pred. Cyclo. Haloperidol Pred. Cyclo

Pred. Cyclo. Azathioprine Pred.

Pred.

Pred. Valproate

Pred. Cyclo. Valproate Methyl Pred. CBZ Pred. Cyclo

Pred. Cyclo

Treatment

Atropy, basal ganglia

Normal

ND

Normal

Cerebral atrophy

NA

Atrophy, posterior infarcts ND

Atrophy, multiple infarcts ND

Cerebral atrophy

Cerebral atrophy

CT scan

No CNS deficit N. school No CNS deficit N. school Slow learner N. school

No CNS deficit N. school Slow learner, epilepsy No CNS deficit N. school No CNS deficit N. school R. Hemiplegia, slow learner, behavioural problems No CNS deficit N. school No CNS deficit N. school No CNS deficit N. school Mild spasticity both LL, dextrusor instability No CNS deficit N. school No CNS deficit N. school Slow learner N. school

Loss to follow up

No CNS deficit N. school Cognitive, CPE

Outcome

230 I. Haji Muhammad Ismail Hussain et al. / Brain & Development 21 (1999) 229–235

11 (4)

10 (13)

10 (37)

11 (5)

10 (4)

11 (2)

19

20

21

22

23

24

F

F

F

F

F

F

Sex

Malay

Malay

Malay

Malay

Malay

Malay

Race Psychosis, dystonic posture, UMN signs Gen. tonic clonic fit Delirium, Gen. fits, transient L. lower limb weakness Gen. fits, cog-nitive/ mood Choreoathetosis/ weakness Gen. fits, aphasia, psychosis

CNS manifestations

Bil. epileptic discharges

Focal sharps/atten uation Normal

Focal sharps

Diffuse slow

Normal

Electroencephalogram

ND

ND

ND

ND

ND

ND

Nerve conduction

Atrophy

Atrophy

Atrophy

Multiple infarcts

Normal

MRI normal

CT scan

Pred. Haloperidol Methyl Pred. Cyclo.

Pred.

Pred.

Pred.

Methyl Pred.

Treatment

Slow learner N. school No CNS deficit N. school No CNS deficit N. school

Died of massive intracranial bleed Epilepsy optic atrophy

No CNS deficit N. school

Outcome

d, duration in months; F, female; M, male; R., right; L., left; Gen., generalised; Bil., bilateral; ND, not done; dcv, decreased conduction velocity; amp., amplitude; Pred., prednisolone; Cyclo, cyclophosphamide; UMN, upper motor neuron; CNS, central nervous system; CPE, complex partial epilepsy; LL, lower limbs; N., normal; CT, computed tomograms; NA, not available.

Age (d)

No.

)deunitnoc(

Table 1

I. Haji Muhammad Ismail Hussain et al. / Brain & Development 21 (1999) 229–235

231

232

I. Haji Muhammad Ismail Hussain et al. / Brain & Development 21 (1999) 229–235

(MRI)scans, electroencephalograms (EEGs), nerve conduction studies, serological parameters, other clinical manifestations of SLE and treatment. All the children were followed up for a year to document their progress and short term outcome. Children with a past history of central nervous system (CNS) infection, hypertensive encephalopathy and uraemia were excluded. Nerve conduction studies, testing both median, both tibial and both sural nerves were only carried out in patients whose parents gave informed consent to the procedure. These studies were carried out with a Medelac Sapphire produced by Vickers Medical. Behavioural assessment was carried out using a translated version of the Rutter’s Behaviour Questionnaire previously validated on Malaysian children by the department of psychiatry, National University of Malaysia [6]. The EEGs in 1996 were carried out on a 10 channel Sanei machine and the follow up EEGs with a Medelac Profile digital system installed in our Department in 1997. All EEGs were carried out in the awake state, with 3 min of hyperventilation and photic stimulation, except in patients having an acute encephalopathy at the time of the recording. In these children a 20-min sleep recording with photic stimulation was performed. As this was a cross-sectional study the recordings were carried out at different stages of the disease. Only in patients 2, 19, 22, 23 and 24 were recordings done early in the disease before steroids were commenced. The study protocol was approved by the research committee of the Ministry of Health.

3. Results There were 24 children with SLE as defined by the ARA who meet the inclusion criteria. There were 23 girls and only one boy giving a female to male ratio of 23:1.There were also no boys among the children who were excluded. The mean age at recruitment was 11.25 years with a range of 8–14 years. There were 14 Malays, eight Chinese, one Dayak and one Indian. Eighteen children (75%) had clinical and neurophysiological evidence of CNS disease. The findings in all 24 children and their functional outcome at 1 year follow up is summarised in Table 1.The duration of SLE at the time the children were recruited into the study is given beside their ages. Overall 11 children (61.1%) had seizures, five each (27.7%) had encephalopathy, headache and psychosis, four (22.2%) had personality changes, three (16.6%) had a stroke, another three had movement disorders, two of whom had chorea and one dystonia and one child (5.5%) had myelitis. In four patients, numbered as 1, 8, 9, and 15, CNS manifestations were the first sign of SLE. Two had an acute encephalopathy treated initially as an acute encephalitis. In one case, the only boy in the series, proteinuria detected while he was being ventilated for his encephalopathy lead to

the diagnosis of SLE. Incidentally his Mycoplasma titer was elevated at 1:320 at the same time. His twin brother developed SLE after the study but has no CNS involvement to date. The second child presented with stupor and generalised seizures without fever. A malar rash was noted on her face leading to the diagnosis. One child presented with hemichorea and was initially diagnosed as Sydenham’s chorea and put on steroids. Anti-streptococcal titres were borderline. She subsequently relapsed and serology confirmed SLE. The last child who presented with CNS case had a stroke and was diagnosed by the collagen screen done to investigate young strokes. Of the 11 children with seizures, nine were generalised, one focal and one had both generalised and focal episodes. All the children with generalised seizures had status epilepticus at some point in their illness. However only three children needed long term anticonvulsants and were given valproate. Six children complained of headache but in one child there was no other clinical evidence of CNS involvement and her EEG was normal so she was not considered a case of CNS lupus. The pain was often severe and the nature was tension-like or migranous at different times. The pain responded to simple analgesics like paracetamol. Five children had encephalopathy, two of whom were mentioned earlier. Three (patients 5, 15, and 18) presented as with fever and were clinically indistinguishable from acute encephalitis. These three all had long term sequelae. Of the three children with stroke, two had hemiplegia, one of whom had residual weakness while the other was lost to follow up and one lower limb monoparesis which resolved. Psychosis was present in five children, two had visual hallucinations, one had visual and auditory and one had visual and tactile hallucinations. Another child had suicidal thoughts. Behavioural assessment showed that four children developed personality changes and mood swings during the course of the follow up. Twenty-one (87.5%) children had at least one EEG done. Eleven (52.8%) were normal, five (23.8%) had focal sharps, three (14.2%) had slow activity, one (4.7%) had focal sharps mixed with electrodecremental events and one (4.7%) had bilateral discharges with focal slow. Fig. 1 shows the EEG of patient 15 during a CNS relapse characterised by worsening seizure, including complex absences, disinhibited behaviour and cognitive decline. Sixteen children had a CAT scan of which only 15 could be traced. In patients 1, 8, 23, and 24 the CAT scan was done before steroids were started. In the other patients the CAT scans were at different intervals after steroids were commenced. The findings are summarised in Tables 2 and 3. Patient 19 had an MRI during an acute bout of psychosis that was normal. Patient 20 had a normal CAT scan but was admitted during a hypertensive crisis with generalised seizures, fundal haemorrhages and deteriorating conscious level. The family had opted for traditional treatment. A repeat scan

233

I. Haji Muhammad Ismail Hussain et al. / Brain & Development 21 (1999) 229–235

Fig. 1. EEG of patient 15 during CNS relapse showing right focal slow activity.

showed massive intracranial bleed. She died the same night. Of the 17 children who had a nerve conduction study, nine were normal, while eight showed evidence of a mononeuritis multiplex. Interestingly abnormalities were only detected in lower limb nerves and none of these children had clinical evidence of a peripheral neuropathy. Three of the children with abnormal nerve conduction had no evidence of cerebral lupus. The details will be reported separately. Among the 18 children with cerebral lupus, 15 had a positive ANF, seven had anti-double stranded DNA (antiDSDNA), 10 had positive LE cells, 13 had reduced compliments and only four were positive for lupus anticoagulant. Only two of the patients with CNS involvement, patients 19 and 24 did not have renal involvement. Ten (55.5%) had a malar rash, another 10 had haematological involvement, six (33.3%) each had arthritis, carditis and oral ulcers, three (16.6%) had a photosensitive rash and one (5.5%) had a discoid rash. On follow up eight children (44.4%) had made a full CNS recovery. Six or one-third had acquired learning difficulties requiring special educational placement. Four (22.2%) had epilepsy but one had been tailed off her anti-epileptic medication. Two (11.1%) patients (8 and 12)

had motor deficits and one (5.5%) had optic atrophy. One child, patient 3, was lost to follow up. Our only death was from hypertensive encephalopathy. As the numbers were too small and many patients had heterogenous CNS involvement it was not possible to analyse the relationship, if any, between presentation and outcome. However, the following observations were made: three of the five children with encephalopathy had residual deficits compared to none of the three children who presented with movement disorders. A normal CAT was a good prognostic factor. The only child with basal ganglia calcification only had mild spastic paraparesis related to her Table 2 Spectrum of manifestations in 18 children with cerebral lupus Neurological manifestation

Number (%)

Seizures Psychosis Encephalopathy Headache Personality changes Stroke Movement disorders Myelitis

11 5 5 5 4 3 3 1

(61) (27.7) (27.7) (27.7) (22.2) (16.6) (16.6) (5.5)

234

I. Haji Muhammad Ismail Hussain et al. / Brain & Development 21 (1999) 229–235

Table 3 Summary of CT scan findings CT scan findings

Number of cases

Cerebral atrophy alone Atrophy with infarcts Multiple infarcts Basal ganglia calcification Normal Not available Total

7 2 2 1 3 1 16

myelitis and is functioning normally. Sharp waves and spikes on EEG even if bilateral did not prelude a bad outcome. However, slow wave activity was associated with cognitive deficits on follow up. For treatment, all the children received prednisolone. Three patients 3, 19, and 24 received methyl prednisolone for their acute CNS complications, and showed improvement within 48 h of starting parenteral steroids. The three children with chorea and dystonia were in addition given haloperidol. The majority of the cases were started on cyclophosphamide for renal indications except for patients 15 and 24 who were started on this drug because of their CNS relapse. None of the other 16 children had a CNS relapse during the period of the study. However, patient 9 had had two episodes of hemichorea in the past but was started on cyclophosphamide for renal relapse. Hence of the 24 children, 12 (including patient 15) had renal relapses, compared to only three ever having a CNS relapse. However, the period of follow up is too short to draw any inference from this.

4. Discussion In our study 18 out of 24 children with SLE or 75% had cerebral lupus. In two previous paediatric series [4, 5] the percentage was 43–44%. The mean age in our children was 11.5 years which is comparable to the previous series namely 12 [4] and 13.3 [5] years, respectively. We found seizures to be the most frequent CNS manifestation, followed by encephalopathy, psychosis and headache. One paediatric series of 40 patients with cerebral lupus found neuropsychiatric changes, namely psychosis, depression, and memory changes in 48% of cases compared to only 20% with seizures. None of our patients had a cranial nerve palsy, noted in both the series. In the majority of paediatric series reviewed by Yancey et al. [4], seizures appear to be the most prominent CNS manifestation. Another point of note is that in only four of our 18 patients was cerebral lupus the presenting complaint leading to the diagnosis of SLE. In the other series 19 of 91 [5] and 13 of 37 [4] children respectively presented with cerebral lupus. This and the overall frequency of cerebral lupus may

represent true population differences, but the number of cases in all the series are small. Although cerebral lupus is common, other causes of neurological deterioration, especially infection, hypertensive and uraemic encephalopathy must be excluded. Infection in particular may be missed in immunosuppressed patients. One of our patients died of cerebral haemorrhage due to hypertensive encephalopathy. Three patients were excluded because of previous CNS infection. All children with a history of renal dialysis for lupus nephritis were also excluded. Cerebral lupus remains largely a clinical diagnosis, and there is no laboratory test to diagnose to confirm or refute it. Most studies have included EEG and neuroimaging to support the diagnosis of CNS lupus. Twenty-one of our patients had an EEG. In one child who only had headache (patient 6), this lead to her being classified as a case of cerebral lupus. However, six children with overt CNS involvement, had normal EEG as shown in Table 1. Other workers have used quantitative EEG to diagnose cerebral lupus and found it to be more sensitive than MRI and CT scanning. However, there was only a 75% specificity with abnormal findings in 28% of patients with no evidence of CNS involvement [7]. All but one of our patients had neuroimaging. Three of 15 CT scans were normal. Of the seven scans that showed atrophy four were from patients that had not yet been started on steroids. This gives credence to cerebral lupus per se being a cause of cerebral atrophy. Such an observation has been previously reported [8]. Only one child had a MRI scan, which was normal. Generally MRI scans are held to be more sensitive than CT [3,9]. Apart from picking up atrophy missed by CT, MRI imaging can also detect focal oedema that resolves with treatment and hence can be used to monitor the efficacy of intervention [10,11]. However, MRI can be normal in established cerebral lupus [3]. As eluded to earlier there are two types of cerebral lupus, diffuse and focal. The exact pathogenesis of these two varieties is still uncertain. In diffuse cerebral lupus with encephalopathy, psychosis and generalised seizures, many of the changes are reversible. These are thought to be due to alterations in electrical impulses and neurotransmitter levels [12]. Antineuronal antibodies levels are elevated in patients with cognitive changes [13]. There also appears to be some specificity in the effects of these antibodies and it is postulated by some that diffuse and focal disease may be due to a variety of antineuronal antibodies [13]. It has been shown that anti-ribosomal P antibody is elevated in lupus psychosis and depression [14]. Focal disease such as strokes may be due to vascular occlusion leading to cell necrosis and microinfarcts [12]. This was thought to be due to vasculitis, but autopsy studies in adult patients failed to show evidence of vasculitis and emboli possibly from Libmans–Sacks endocarditis may be more important than previously thought [15]. Others have proposed a non-inflammatory vasculopathy with thickening

I. Haji Muhammad Ismail Hussain et al. / Brain & Development 21 (1999) 229–235

of the vascular wall, intimal proliferation and microinfarcts [12]. Furthermore, antiphospholipid antibodies that bind to endothelial cells and platelet membranes may result in thrombosis aggravating the vascular injury [12]. The concept of hypercoagulability in cerebral lupus has important implications for treatment. Anticoagulation may be just as important as immunosuppression. In our series only four patients had lupus anticoagulant, however, we did not study all cases during the acute phase nor did we measure anticardiolipin antibodies. The finding of low compliments in 13 of our 18 cerebral lupus cases would in fact be in keeping with the concept that recurrent low grade vasculitis is the underlying basis for the vasculopathy [3]. A recent study using serial SPECT in four children with cerebral lupus documented improvement in perfusion even in a child who presented with right hemiparesis and numbness. Her first scan showed hypoperfusion in the left temporal and temporoparietal regions, which normalised 3 weeks after treatment was started. This supports a reversible pathogenetic mechanism for even the focal forms of cerebral lupus [16]. There is no randomised controlled trials on the treatment of cerebral lupus and we actually followed the regime used by the nephrologist in our Institute. We used methylprednisolone at a dose of 30 mg/kg per day for 3 days followed by 6 weeks of oral steroids at 2 mg/kg. Improvement was seen by the third day of methylpredisolone. The two patients on cyclophosphamide are on monthly pulses, for 6 months. In our small series only 44% made a full recovery. This compares poorly with the figure of 90% and 78%, respectively in the series of Steinlin et al. [5] and Yancey et al. [4]. The majority of deficits in our series were cognitive, affecting one-third of our children who required remedial education. Only four of 11 with seizures needed long term anticonvulsants, the others became fit free with immunosuppressive therapy. Sixteen of our 18 children with cerebral lupus had concurrent renal lupus and their overall prognosis is guarded. In conclusion, cerebral lupus is a common complication of childhood SLE, affecting 75% of patients in our series. The diagnosis is essentially clinical as there are no markers specific for CNS lupus. Changes are often detected in neuroimaging studies and EEG, some of which can be used to monitor the effects of treatment. The actual pathogenesis of this condition is still uncertain and is probably multifactorial involving immune, inflammatory and thrombotic pathways. At least some of these changes are reversible and all these children should receive aggressive immunosuppression. The role of anticoagulants in treatment and prophylaxis is uncertain. The overall outlook for functional recovery is good, but long term prognosis probably depends on the associated systemic disease, especially renal involvement.

235

Acknowledgements We would like to thank the Director-General of Health, Malaysia for permission to publish this article.

References [1] Lehman TJ. A practical guide to systemic lupus erythematosus. Pediatr Clin North Am 1995;42:1223–1238. [2] Tan EM, Cohen AS, Fries JF, Masi AT, McShane DJ. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthrit Rheum 1982;25:1271–1277. [3] Mitchell I, Webb M, Hughes R, Stewart J. Cerebral lupus. Lancet 1994;343:579–582. [4] Yancey CL, Doughty RA, Athreya BH. Central nervous system involvement in childhood systemic lupus erythematosus. Arthrit Rheum 1981;24:1389–1395. [5] Steinlin MI, Blaser SI, Gilday DL, Eddy AA, Logan WJ, Laxer RM. et al. Neurologic manifestations of pediatric systemic lupus erythematosus. Pediatr Neurol 1995;13:191–197. [6] Cheah YC, Kasmini K, Amin SM, Zasmani S. Validation of Rutter’s Behaviour Questionnaire for Malaysian Children. Malaysian J Psychiatry 1995;3:65–73. [7] Ritchlin CT, Chabot RJ, Alper K, Buyon J, Belmont HM, Roubey R. et al. Quantitative electroencephalography. A new approach to the diagnosis of cerebral dysfunction in systemic lupus erythematosus. Arthrit Rheum 1992;35:1330–1342. [8] Raymond AA, Zariah AA, Samad SA, Chin CN, Kong NC. Brain calcification in patients with cerebral lupus. Lupus 1996;5:123– 128. [9] Sibbitt WL Jr, Sibbitt RR, Griffey RH, Eckel C, Bankhurst AD. Magnetic reasonance and computed tomographic imaging in the evaluation of acute neuropsychiatric disease in systemic lupus erythematosus. Ann Rheum Dis 1989;48:1014–1022. [10] 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. [11] Russo R, Gilday D, Laxer RM, Eddy A, Silverman ED. Single photon emission computed tomography scanning in childhood systemic lupus erythematosus. J Rheumatol 1998;25:576–582. [12] West SG. Neuropsychiatric lupus. Rheum Dis Clin North Am 1994;20:129–158. [13] Kelly MC, Denburg JA. Cerebrospinal fluid immunoglobulins and neuronal antibodies in neuropsychiatric systemic lupus erythematosus and related conditions. J Rheumatol 1987;14:740–744. [14] Watanabe T, Sato T, Uchiumi T, Arakawa M. Neuropsychiatric manifestations in patients with systemic lupus erythematosus: diagnostic and predictive value of longitudinal examination of anti-ribosomal P antibody. Lupus 1996;5:178–183. [15] Devinsky O, Petito CK, Alonso DR. Clinical and neuropathological findings in systemic lupus erythematosus: the role of vasculitis, heart emboli, and thrombotic thrombocytopenic purpura. Ann Neurol 1988;23:380–384. [16] Huang JL, Yeh KW, You DL, Hsieh KH. Serial single photon emission computed tomography imaging in patients with cerebral lupus during acute exacerbation and after treatment. Pediatric Neurol 1997;17:44–48.