The prevalence and clinical characteristics of systemic lupus erythematosuswith infectious brain lesions in China

Y Xu*, D Xu*, T Zhang, X-M Leng, F-C Zhang, X-F Zeng

Department of Rheumatology, Peking Union Medical College Hospital, Peking Union Medical College, Chinese Academy of MedicalScience, Beijing, China

Objective: Infectious brain lesions (IBLs) are life-threatening in patients with systemic lupus erythematosus (SLE). Theaim of this study was to determine the prevalence of IBL in SLE patients and the clinical characteristics of SLE patientswith IBL.Methods: Medical charts of 15 consecutive SLE patients with IBL admitted to Peking Union Medical College Hospital(PUMCH) from January 1995 to October 2010 were reviewed systematically. A total of 150 cases were randomly selectedas controls from 4115 SLE inpatients without IBL in PUMCH during the same period.Results: The prevalence of IBL in SLE patients was 0.4%. Significant differences were observed between SLE patientswith and without IBL in the following manifestations (p < 0.05): arthritis/musculoskeletal involvement (66.7% vs.32.0%), C-reactive protein (CRP) elevation (84.6% vs. 28.0%), anti-dsDNA antibody positivity (13.3% vs. 42.9%), andelevated SLE Disease Activity Index (SLEDAI) score (> 5) (13.3% vs. 71.3%). Fever was the most commonmanifestation (80%), followed by headache and focal neurological signs (73.3%). Twelve patients presented withinfections in other sites, including pulmonary infection (66.7%) and meningitis (40.0%). Enhanced cranial magneticresonance imaging (MRI) revealed point-enhancing or ring-enhancing lesions in all patients evaluated (12/12, 100%).Mycobacterium tuberculosis was the most common pathogen (10 cases, 66.7%). After administration of antibioticstargeting the pathogens, 11 patients (73.3%) recovered.Conclusions: IBL is not common in SLE patients. In stable SLE patients with fever, focal neurological signs, and CRPelevation, IBL should be suspected. Enhanced cranial MRI and a thorough check-up should be performed in a timelymanner. It is very important to identify the pathogens and initiate treatment as early as possible.

Systemic lupus erythematosus (SLE) is a multisystemicautoimmune disease. SLE patients often suffer fromimmunological dysfunction resulting from the diseaseitself, including complement deficiencies, mannose-binding lectin, polymorphisms, elevated Fcgamma IIIand granulocyte–macrophage colony-stimulating factor(GM-CSF) levels (1). Furthermore, immunosuppressivetherapy, especially with glucocorticoids, usually used inthe treatment of moderate and severe lupus, increases therisk of infections in SLE patients. Therefore, prednisonetreatment, even at moderate doses, is one of the signifi-cant and independent predictors of major infections inSLE patients (2). Thus both immunological aberrations

of SLE and immunosuppressive agents contribute to thesusceptibility of SLE patients to infections. These caninclude viral, bacterial, fungal, and parasitic infections,involving multiple organs (1). Infections are an impor-tant cause of mortality in SLE patients. Identifying theinfections in SLE patients remains difficult because ofatypical infectious manifestations that can sometimesmimic lupus. Central nervous system (CNS) infectionsin SLE patients, such as meningitis and infectious brainlesions (IBLs), are life-threatening and account for asignificantly increased mortality. IBLs, which includeall infectious space-occupying lesions in the brainparenchyma, are rare in SLE patients (3). These lesionsdevelop when pathogenic micro-organisms reach thebrain parenchyma through haematogenous dissemina-tion. Although IBLs show manifestations similar toneuropsychiatric SLE (NPSLE), the treatment strategiesare markedly different, underlying the need for adequatedistinction and diagnosis. An accurate diagnosis and theinitiation of appropriate treatment as early as possibleare crucial (4). To clarify the prevalence and clinicalcharacteristics of IBL in SLE patients and to direct

DongXu, Department of Rheumatology, Peking UnionMedical CollegeHospital, Peking Union Medical College, Chinese Academy of MedicalScience, 41 Damucang Hutong, Xicheng District, 100032 Beijing,China.Email: xudong74@hotmail.com

* Both authors contributed equally to this work.

Accepted 26 March 2012

466 Scand J Rheumatol 2012;41:466–471

© 2012 Taylor & Francis on license from Scandinavian Rheumatology Research Foundation

DOI: 10.3109/03009742.2012.680607www.scandjrheumatol.dk

Scan

d J

Rhe

umat

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

aure

ntia

n U

nive

rsity

on

10/0

1/13

For

pers

onal

use

onl

y.

further clinical management, we retrospectively ana-lysed the characteristics of a patient group in China.

Patients and methods

Patients

Medical charts of 15 consecutive SLE patients with IBLadmitted to Peking Union Medical College Hospital(PUMCH) from January 1995 to October 2010 were sys-tematically reviewed: characteristics studied includeddemographic data, clinical features, laboratory findings,treatments, and outcomes. A total of 150 cases wererandomly selected as controls from 4115 SLE inpatientswithout IBL in PUMCH during the same period. Allpatients fulfilled the 1997 American College ofRheumatology revised classification criteria forSLE. Active SLE was defined as an SLE DiseaseActivity Index (SLEDAI) score > 5. A definitive diagno-sis of IBL was based on clinical manifestations, cranialimaging strongly suggestive of IBL, and one of the fol-lowing criteria: a positive result for anymicro-organism inbrain tissue specimens, cerebrospinal fluid (CSF), bloodor extracranial tissue specimens through culture, Gramstain, or acid-fast staining; typical imaging suggestive ofextracranial infections, such as chest computed tomogra-phy (CT), signifying miliary pulmonary tuberculosis; oreffectiveness achieved with empirical treatment.

Statistical analyses

SPSS version 11.0 (SPSS, Chicago, IL, USA) was usedfor data processing and analysis. Continuous variables(mean � SD) were determined using the non-parametrictest. Categorical variables were determined usingPearson’s χ2 test (using the continuity corrected χ2 whenthe minimum expected count was < 5 and using Fisher’sexact test when the minimum expected count was < 1).In all tests, p-values < 0.05 were considered significant.

Results

Clinical characteristics and outcomes

The overall prevalence of IBL in SLE patients was 0.4%(15 out of 4115 cases). The duration from onset of SLE toIBL was 7.8 � 9.8 (range 0.6–30.0) years. Univariateanalysis showed that, compared to SLE patients withoutIBL, those with IBL were less active and had more arthri-tis/musculoskeletal involvement (Table 1).

All patients had received high doses of glucocorticoid(1 mg/kg/day equivalent dose of prednisone), with fourundergoing methylprednisolone pulse therapy.Immunosuppressants, including cyclophosphamide (11cases), methotrexate (one case), hydroxychloroquine(two cases), and glucosida Tripterygii TOTA (twocases), were administered in 14 patients (93.3%) discon-tinuously. When IBL was diagnosed, all patients weremanaged with prednisone daily (26.5 � 16.8 mg/day,range 5–60 mg/day), and six patients (40%) were man-aged with immunosuppressants (five with cyclophos-phamide and one with glucosida Tripterygii TOTA).

The age at onset of IBL was 38.0� 12.1 (range 20–59)years. The disease course of IBL was 2.0 � 1.9 (range0.1–7.0) months. The infectious manifestations includedfever (12 cases, 80%), headache (11 cases, 73.3%), nau-sea/vomiting (eight cases, 53.3%), focal neurologicalsigns (11 cases, 73.3%), consciousness disorders (threecases, 20.0%), convulsion (two cases, 13.3%), productivecough (three cases, 20.0%), and night sweating (twocases, 13.3%). Twelve cases (80%) showed infections inother sites (Table 2). Of note, only three out of 10 patientswith pulmonary infection had respiratory manifestations.After targeted antibiotics treatment and appropriate surgi-cal drainage for severe complications, 11 patients (73.3%)recovered, two patients (13.3%, one tuberculosis (TB)and one nocardiosis) died, and two patients (13.3%) failedto follow up. The mortality rate was 13.3%. Detailedinformation of pathogens, treatment, and outcome isshown in Table 2.

Table 1. Clinical characteristics of SLE patients with and without IBL.

SLE with IBL SLE withoutI BL(n ¼ 15) (n ¼ 150) p value

Age of SLE onset (years) 30.2 � 10.6 32 � 13.6 0.727Sex ratio (men/women) 2/13 22/128 1.00Elevation of SLEDAI score (> 5) 2 (13.3) 107 (71.3) < 0.001Mucocutaneous involvement 12 (80.0) 71 (47.3) 0.117Renal involvement 11 (73.3) 93 (62.0) 0.386Nephrotic syndrome 3 (20) 53 (35.3) 0.232

Arthritis/musculoskeletal involvement 10 (66.7) 48 (32.0) 0.007Haematological system involvement 8 (53.3) 60 (40.0) 0.317Nervous system involvement 5 (33.3) 30 (20.0) 0.383Central nervous system 4 (26.7) 25 (16.7) 0.539

Respiratory system involvement 2 (13.3) 26 (17.3) 0.974Administration of cyclophosphamide 11 (73.3) 109 (72.7) 1.000Administration of hydroxychloroquine 2 (13.3) 16 (10.7) 1.000

Data are listed as mean � SD or n (%).

Infectious brain lesions in SLE 467

www.scandjrheumatol.dk

Scan

d J

Rhe

umat

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

aure

ntia

n U

nive

rsity

on

10/0

1/13

For

pers

onal

use

onl

y.

Table2.

Charac

teristicsof

infections

inSLEpa

tientswith

IBL.

Aetiologica

ldiagn

osis

Historyof

infection

Infectionsites

othe

rtha

nbrainpa

renc

hyma

ChestC

TRe

source

Pathog

enDiag

nosis

Trea

tmen

tOu

tcom

e

1N

Leg

NAb

scessinleg

Streptothrix

noca

rdii

Streptothrixno

cardii(legab

scessan

dbrain

abscess)

Merop

enem

andsurgical

draina

geRe

covered

2N

Forearm,sho

ulde

rN

Subc

utan

eous

abscess

Noc

ardia

asteroides

Noc

ardiaasteroides

(sub

cutane

ousab

scess

andbrainab

scess)

TMPc

oRe

covered

3N

NN

Cerebrallesion

sNoc

ardia

brasilien

sis

Noc

ardiabrasilien

sis(brainab

scess)

Gentam

icinþ

ciprofloxacin

þmerop

enem

andsurgical

draina

ge

Death

4TB

Lung

Leftup

perinfiltrate

NN

Unde

term

ined

bacterialinfec

tion(brain

abscess)

Merop

enem

Reco

vered

5TB

Lung

Miliaryno

dules

Sputum

NTM

NTM

(brainab

scess)

Clarithromycinþ

levofloxacin

þEM

BRe

covered

6N

Men

inge

sN

Cerebrallesion

sTB

TB(cereb

raltub

ercu

loma)

INHþ

RFPþ

EMBþ

PZA

Reco

vered

7N

Scalp,thigh

NAb

scessesinscalp

andthigh

TBTB

(sca

lp,thigh

,and

brainab

scesses)

INHþ

RFPþ

EMBþ

PZA

Reco

vered

8TB

Men

inge

s,lung

Infiltra

tean

dno

dulesof

right

uppe

rlob

e

NN

TB(cereb

raltub

ercu

lomaan

dmen

ingitis

)INHþ

RFPþ

EMBþ

PZA

Reco

vered

9TB

Men

inge

s,lung

Miliaryno

dules

NN

TB(cereb

raltub

ercu

loma,men

ingitis,and

pulmon

arytube

rculosis)

INHþ

RFPþ

EMBþ

PZAþ

SMMissed

10N

Smallintestine,lung

Miliaryno

dules

Abscessofintestine

TBTB

(cereb

raltub

ercu

loma,pu

lmon

aryan

dintestinaltube

rculosis)

INHþ

RFPþ

EMBþ

PZA

Missed

11N

Men

inge

s,lung

Leftup

perlob

einfiltra

teCS

FTB

TB(cereb

raltub

ercu

lomaan

dmen

ingitis)

INHþ

EMBþ

PZA

þrifap

entineþ

levofloxacin

Reco

vered

12N

Men

inge

s,lung

Miliaryno

dules

NN

TB(cereb

raltub

ercu

lomaan

dmen

ingitis)

INHþ

RFPþ

EMBþ

PZA

Reco

vered

13N

Hand

,kne

e,lung

Righ

tupp

erlobe

infiltra

teBloo

darticular

abscess

TBTB

(cereb

raltub

ercu

lomaan

darticular

abscess)

INHþ

RFPþ

EMBþ

PZA

Death

14N

NN

NN

TB(cereb

raltub

ercu

loma)

INHþ

RFPþ

PZAþ

amikac

inþ

levofloxacin

Reco

vered

15N

Men

inge

s,lung

Miliaryno

dules

CSF,bloo

dTB

TB(cereb

raltub

ercu

loma,men

ingitis,and

pulmon

arytube

rculosis)

INHþ

RFPþ

PZAþ

amikac

inþ

levofloxacin

Reco

vered

SLE,

System

iclupu

serythe

matosus;IBL

,infec

tious

brainlesion

s;CT

,com

putedtomog

raph

y;N,N

egative;

CSF,

cerebrospina

lfluid;N

TM,n

on-tu

berculou

smycob

acteria

;TB,

tube

rculosis;

INH,

ison

iazid

;RFP,rifampicin;EM

B,etha

mbu

tol;PZ

A,pyrazin

amide;TM

Pco,sulfametho

xazoleþ

trimetho

prim.

468 Y Xu et al

www.scandjrheumatol.dk

Scan

d J

Rhe

umat

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

aure

ntia

n U

nive

rsity

on

10/0

1/13

For

pers

onal

use

onl

y.

Laboratory characteristics

Univariate analysis showed that compared to SLEpatients without IBL, those with IBL had a higher inci-dence of C-reactive protein (CRP) elevation (84.6% vs.28.0%, p < 0.001) and a lower anti-dsDNA antibodypositive rate (13.3% vs. 42.9%, p ¼ 0.026).Leucopaenia was identified in five cases (33.3%), neu-

tropaenia in two (13.3%), and lymphopaenia in 12 (80.0%).Lymphocyte subset analysis was performed in sevenpatients. A depressed B-cell function was observed inseven cases (100%) and a depressed T-cell function infive (71.4%). CD4þ T-cell count was decreased in sixcases and was < 200 cells/mm3 in four cases.Hypoalbuminaemia was observed in six cases (6/14,42.9%).Lumbar puncture revealed that CSF opening pressure

was 251.8 � 68.6 (range 180–350) cmH2O. CSF proteinlevel in these patients was 0.96 � 0.66 (range 0.16–2.34)g/L and elevated protein levels were found in 11 cases(73.3%). However, increased white blood cell count inCSF was only observed in three cases (20%). Decreasedglucose in CSF was observed in five cases (33.3%).Cranial imaging examinations were performed for all

patients, including CT scanning for one patient, magneticresonance imaging (MRI) for two patients, and enhancedMRI for 12 patients. All examinations showed evidenceof space-occupying cranial lesions. Single lesions werefound in four cases (26.7%) and multiple lesions in 11(73.3%). Severe perifocal oedema was observed in sixcases (40.0%) and midline shift in five (33.3%). Allpatients examined by cranial MRI showed point- orring-enhancing lesions (12/12, 100%). IBL was hypoin-tense or isointense in T1-weighted images (T1WI) andhyperintense in T2-weighted images (T2WI). Threepatients underwent diffusion-weighted MR imaging(DWI), in which IBL was hyperintense (100%)(Figure 1).

Discussion

Yang et al (5) previously reported that hypoalbuminaemiaand administration of large doses of prednisone at theonset of CNS infection or higher mean doses of predni-sone within the previous year were important risk factorsfor the development of CNS infections in SLE patients. Inthe current study the patients were administered highdoses of glucocorticoid because of moderate and severelupus manifestations. However, some studies have shownthat the use of cyclophosphamide increases the risk ofinfection in SLE patients (6), while antimalarials have aprotective effect (2), Although there were no significantdifferences between patients with and without IBL in theuse of antimalarials, no patient in the current study wasusing antimalarials prior to or upon infection. To decreasethe risk for infections, prednisone should be used withcaution, and maintenance doses of > 5 mg/day are bestavoided. In addition, treatment with hydroxychloroquinewould be indicated in all patients with SLE without con-traindications (2). Lymphopaenia and depressed B- andT-cell function indicated suppressed humoral and cellularimmunity in these SLE patients. Therefore, it can beassumed that the immunological dysfunction induced bythe long course of SLE itself and the treatments for SLEmay be responsible for development of IBL. Meanwhile,hypoalbuminaemia was observed in 42.9% patients, sug-gesting that malnutrition may also contribute to IBL.

When neuropsychiatric manifestations (such as head-ache and seizures) are present, NPSLE should be conside-red, especially in the absence of fever. However, beforeconfirming NPSLE it is necessary to exclude CNS infec-tions because of the distinct treatment strategy used foreach. IBL, which could mimic the presentation ofNPSLE, should not be ruled out, especially when SLE isstable. It is generally accepted that patients with NPSLEshow high disease activity (7). Our data showed thatdisease activity is lower in patients with IBL. Moreover,

A B C

Figure 1. Cranial MRI of SLE patients with IBL (A) T1-Weighted image, (B) T2-Weigthed image, and (C) diffusion-weighted image.

Infectious brain lesions in SLE 469

www.scandjrheumatol.dk

Scan

d J

Rhe

umat

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

aure

ntia

n U

nive

rsity

on

10/0

1/13

For

pers

onal

use

onl

y.

in contrast to other infections, in which fever is the mostcommon clinical feature, 20% patients with IBL in ourstudy did not suffer from fever. This further increased thechallenge of differentiating IBL from NPSLE. We alsofound that arthritis/musculoskeletal involvement andCRP elevation were more frequent in patients withIBL. CRP is an acute-phase reactant that is found to beelevated in infections and many autoimmune states. CRPelevation is more common in infectious patients than inactive SLE patients and does not reflect SLE diseaseactivity (1). Kim et al recently demonstrated that CRPlevels of SLE patients with infections were higher thanthose with flares (5.9 vs. 0.06 mg/dL, p < 0.001) (8).Therefore, CRP levels can help to distinguish infectionsfrom lupus flare, especially when the CRP levels are > 50mg/L. Infections could induce and/or aggravate joint andmuscular symptoms, which may explain why arthritis/musculoskeletal involvement was more likely to be pre-sent in SLE patients with IBL. However, unless micro-organisms are found in articular cavity or muscular tissue,it is very difficult to confirm whether infection or SLEinduces the arthritis/musculoskeletal involvement.We found that CSF examination in the diagnosis of IBL

was not as helpful as in the diagnosis of infectious menin-gitis. However, when patients have infectious meningitissimultaneously, a CSF examination may be useful to iden-tify the pathogens. Furthermore, Hopia et al reportedincreased protein levels of a proliferation-inducing ligand(APRIL), a type of B-cell stimulation and survival mole-cule, in the CSF from patients with NPSLE compared toother neurological diseases (9). However, whether APRILcould become a useful tool for differentiating IBL fromNPSLE is not known.Cranial imaging, especially enhanced MRI, is impor-

tant for the diagnosis of IBL. Point- or ring-enhancinglesions with perifocal oedema or midline shift are rela-tively specific for IBL. Moreover, DWI could differenti-ate IBL (hyperintense in DWI and hypointense inapparent diffusion coefficient, ADC) from a neoplasticlesion (hypointense or variably hyperintense in DWI andhyperintense in ADC) (10).Identification of pathogens is crucial for further treat-

ment. Because biopsy of a brain lesion brings great riskfor some patients, leading to hemiplegia and even death insome cases, greater emphasis should be placed onexamining other potential infection sites for pathogen iden-tification (11). Our data showed that most of our patientshad infections in other sites, even without relevant symp-toms. Therefore, a thorough check-up, especially a pulmon-ary examination, was important for further treatment. It hasbeen reported thatMycobacterium tuberculosis (TB), non-tuberculous mycobacteria (NTM), bacterial species, andfungi have been causes of IBL in SLE patients (3, 5, 12).SLE patients have an increased susceptibility to TB (13)and TB is the most common pathogen of CNS infection inSLE patients (5), with meningitis as the most frequentpattern, followed by TB brain abscess or tuberculoma(14). Common bacteria cultured from brain abscesses

include Viridans streptococci, Staphylococcus aureus,Bacteroides fragilis group/species, and anaerobic strepto-cocci (15, 16). Opportunistic bacteria, such as Listeria andNocardia, could cause a brain abscess in immunocompro-mised hosts (17, 18). Mok et al reported that the lung wasthe most common location of Nocardia infection (81%),followed by the CNS (13%), with high mortality (75%)(19). In our study, 66.7% were identified as TB, followedby Nocardia (20%).

The published mortality of intracranial tuberculoma orbrain abscess varies in different studies, and death isusually caused by a delay in diagnosis and treatment (4,14, 15, 20). Therefore, early diagnoses and treatment areimportant. If pathogens cannot be identified, empiricalantimicrobial therapy should be initiated according tothe common pathogens in our data.

In conclusion, IBL should be highly suspected in SLEpatients with unexplained fever, headache, focal neurolo-gical signs, and CPR elevation, especially when SLE isstable, and cranial enhanced MRI should be performedquickly. A careful check-up, especially pulmonaryexamination, should be considered when the pathogenis difficult to identify, even if no related symptomsexist. TB is the most common pathogen inIBL. Therefore, administration of antibiotics targetingthe pathogens or empirical antibiotics at an early stagecould improve the outcome.

References1. Zandman-Goddard G, Shoenfeld Y. Infections and SLE.

Autoimmunity 2005;38:473–85.2. Ruiz-Irastorza G, Olivares N, Ruiz-Arruza I, Martinez-Berriotxoa

A, Egurbide MV, Aguirre C. Predictors of major infections insystemic lupus erythematosus. Arthritis Res Ther 2009;11:R109.

3. Vargas PJ, King G, Navarra SV. Central nervous system infectionsin Filipino patients with systemic lupus erythematosus. Int J RheumDis 2009;12:234–8.

4. Tseng JH, Tseng MY. Brain abscess in 142 patients: factorsinfluencing outcome and mortality. Surg Neurol 2006;65:557–62.

5. Yang CD, Wang XD, Ye S, Gu YY, Bao CD, Wang Y, et al.Clinical features, prognostic and risk factors of central nervoussystem infections in patients with systemic lupus erythematosus.Clin Rheumatol 2007;26:895–901.

6. Noel V, Lortholary O, Casassus P, Cohen P, Genereau T, AndreMH, et al. Risk factors and prognostic influence of infection in asingle cohort of 87 adults with systemic lupus erythematosus. AnnRheum Dis 2001;60:1141–4.

7. Unterman A, Nolte JE, Boaz M, Abady M, Shoenfeld Y, Zandman-Goddard G. Neuropsychiatric syndromes in systemic lupus erythe-matosus: a metaanalysis. Semin Arthritis Rheum 2010; 41:1–11.

8. Kim HA, Jeon JY, An JM, Koh BR, Suh CH. C-reactive protein is amore sensitive and specific marker for diagnosing bacterialinfections in systemic lupus erythematosus compared to S100A8/A9 and procalcitonin. J Rheumatol 2012;39:728–34.

9. Hopia L, Thangarajh M, Khademi M, Laveskog A, Wallstrom E,Svenungsson E, et al. Cerebrospinal fluid levels of a proliferation-inducing ligand (APRIL) are increased in patients with neurop-sychiatric systemic lupus erythematosus. Scand J Rheumatol 2011;40:363–72.

10. Chang SC, Lai PH, Chen WL, Weng HH, Ho JT, Wang JS, et al.Diffusion-weighted MRI features of brain abscess and cystic ornecrotic brain tumors: comparison with conventional MRI. ClinImaging 2002;26:227–36.

470 Y Xu et al

www.scandjrheumatol.dk

Scan

d J

Rhe

umat

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

aure

ntia

n U

nive

rsity

on

10/0

1/13

For

pers

onal

use

onl

y.

11. Cherian A, Thomas SV. Central nervous system tuberculosis. AfrHealth Sci 2011;11:116–27.

12. MokMY,Wong SS, Chan TM, FongDY,WongWS, Lau CS. Non-tuberculous mycobacterial infection in patients with systemic lupuserythematosus. Rheumatology (Oxford) 2007; 46:280–4.

13. Prabu V, Agrawal S. Systemic lupus erythematosus and tuber-culosis: a review of complex interactions of complicated diseases.J Postgrad Med 2010;56:244–50.

14. Arvanitakis Z, Long RL, Hershfield ES, Manfreda J, Kabani A,Kunimoto D, et al. Tuberculosis molecular variation in CNS in-fection: evidence for strain-dependent neurovirulence. Neurology1998;50:1827–32.

15. Prasad KN, Mishra AM, Gupta D, Husain N, Husain M, Gupta RK.Analysis of microbial etiology and mortality in patients with brainabscess. J Infect 2006;53:221–7.

16. Le Moal G, Landron C, Grollier G, Bataille B, Roblot F, Nassans P,et al. Characteristics of brain abscess with isolation of anaerobicbacteria. Scand J Infect Dis 2003;35:318–21.

17. Cone LA, Leung MM, Byrd RG, Annunziata GM, Lam RY, HermanBK. Multiple cerebral abscesses because of listeria monocytogenes:three case reports and a literature reviewof supratentorial listerial brainabscess(es). Surg Neurol 2003; 59: 320–8.

18. Fleetwood IG, Embil JM, Ross IB. Nocardia asteroides cerebralabscess in immunocompetent hosts: report of three cases and reviewof surgical recommendations. Surg Neurol 2000; 53:605–10.

19. Mok CC, Yuen KY, Lau CS. Nocardiosis in systemic lupuserythematosus. Semin Arthritis Rheum 1997;26:675–83.

20. El Sahly HM, Teeter LD, Pan X, Musser JM, Graviss EA. Mortalityassociated with central nervous system tuberculosis. J Infect2007;55:502–9.

Infectious brain lesions in SLE 471

www.scandjrheumatol.dk

Scan

d J

Rhe

umat

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y L

aure

ntia

n U

nive

rsity

on

10/0

1/13

For

pers

onal

use

onl

y.