Diagnosis of spirochetal meningitis by enzyme linked immunosorbent assay and indirect immunofluorescence assay in serum and cerebrospinal fluid

0095-1137/85/050819-07$02.00/0

CopyrightC) 1985,AmericanSociety for Microbiology

Diagnosis of Spirochetal

Meningitis by

Enzyme-Linked

Immunosorbent Assay and

Indirect Immunofluorescence Assay in

Serum and

Cerebrospinal Fluid

GORAN T.

STIERNSTEDT,l*

MARTA GRANSTROM,2 BENGT HEDERSTEDT,2 AND BIRGIT

SKOLDENBERG'

Karolinska Institute, Department of InfectiousDiseases, Danderyd Hospital, S-182 88

Danderyd,1

andDepartmentof

Bacteriology,

National

Bacteriological Laboratory,

S-105 21

Stockholm,2

Sweden

Received 19November1984/Accepted6February 1985

Theantibodyresponse against a spirochetal strain isolated from Swedish Ixodes ricinus ticks was determined byenzyme-linked immunosorbent assay(ELISA)and indirect immunofluorescenceassay ofcerebrospinal fluid (CSF) and serumspecimens from 45 patients with chronic meningitis. Samples of CSF, serum, or both from patients with various infections of the central nervous system, multiple sclerosis, syphilis, or infectious mononucleosis and from healthy individuals served as control samples. Probablespirochetal etiologycouldbe demonstrated for 41 of 45(91%) patientswithclinicalsymptoms of chronic meningitis. Approximately 25% of thepatients had significantly elevated titers of antibody to the spirochete in CSF but notinserum. Thehighest

diagnostic sensitivity, 91 %, was demonstrated by measurement of CSF antibodies and calculation of a spirochetal CSFtiter index, which is the ratio of(ELISA titer in CSF/ELISAtiter in serum) to (albumin in

CSF/albumin in serum) and which also considers the degree of blood-CSF barrier damage. The highest

specificity, 98%, was obtained by calculation of a CSF titer index. Patients with short duration of disease were especially prone to be antibody negative in serum butpositive in CSF. Significantrise in serumantibody titers was seldom demonstrated in patients treated with antibiotics. It is concluded that measurement of CSF antibodies, especially by ELISA, is a highly sensitive and specific method for the immunological diagnosis of spirochetal meningitis.

In Europe, neurological disease with radicular pain and

chroniclymphocytic meningitis after tickbitewas described

in1922byGarinandBujadoux(8)andin 1941by Bannwarth

(3, 4). In a few cases the neurological symptoms were

preceded by erythema chronicummigrans,askin lesion first

described bythe Swedishphysician AlfredAfzelius in 1909

(2). In 1983, we reported persistent or progressive chronic

meningitis (CM) in 21 patients without demonstrable

infec-tionsorother disease. Thepatientshad intrathecal synthesis

of considerable quantities of immunoglobulin G, and they

improved or recovered during treatment with high-dose

intravenous penicillin (13). We discussed the possible

con-nection withLymedisease, first describedin 1976fromthe

community of Lyme, Connecticut, by Steereet al. (16).

Lyme disease usually starts with erythema chronicum

migrans and may then be followed by neurological and

cardiac as well as arthritic complications (15). Recently,

Burgdorfer and co-workers succeeded in isolating a

spiro-chete from one vector of the disease, the tick, Ixodes dammini (6). This spirochete is now officially named Bor-relia burgdorferi (9). The role of B. burgdorferi as the

etiologicalagentofLyme disease is supported by induction

of a disease resembling erythema chronicum migrans in

rabbits exposed to ticks, by high antibody titers to this

spirochete in patients with Lyme disease (6), and by

culti-vation ofspirochetes from blood, cerebrospinal fluid (CSF), orskin lesions from a few Lyme patients (5, 14). Steere in the United States (14) andAckermann in the Federal

Repub-lic of Germany (1) have examined antibodies to B.

burg-dorferi by indirect immunofluorescence assay (IFA).

Re-cently Russell (11) and Craft (7) published comparative

*Correspondingauthor.

studies between IFA and enzyme-linked immunosorbent

assay(ELISA)performed on sera.To ourknowledge, there have not been published any comparative studies between

IFA and ELISA performed on both serum and CSF in

parallel. We have recently presented preliminary data on

serumantibodiesagainstB. burgdorferi isolated from Amer-ican I. dammini and against a spirochete isolated from

Swedish Ixodes ricinus in sera from patients with CM

measured byIFA(16a). The aimofthepresentinvestigation

is tocompare IFA and ELISA results with measurementsof

antibodies in serum andCSF forthelaboratorydiagnosis of

I. ricinus-transmitted spirochetal infections of the central

nervoussystem (CNS).

MATERIALS AND METHODS

CSF and serumspecimens. Table1shows thenumbersand sourcesofspecimens testedby IFA and ELISA.

Patients with CM. Forty five patients were hospitalized from 1975 to 1983 with CM. CM was defined as a disease which failed to improve or worsen clinically in its CSF abnormalities during at least a 2-week course (13, 16a). Infectious CNS diseases caused by Toxoplasma gondii,

Cryptococcus neoformans, Mycobacterium tuberculosis,

Listeria monocytogenes, and Treponema pallidum were excluded in all 45patients. Forty-three of the 45 patients had been treated with high-dose intravenous antibiotics.

Controls. Asmeningitis controls, CSF and serum samples

from patients with different infectious diseases of the CNS

werestudied. The infections included tick-borne encephali-tis (TBE), varicella-zoster virus, herpes simplex virus,

cytomegalovirus, Epstein-Barr virus,measles or echovirus,

Mycoplasmapneumoniae, and acute bacterial and

tubercu-lous orcryptococcal infections. We also examined CSF and

819

on April 11, 2020 by guest

http://jcm.asm.org/

TABLE 1. Patientsand control subjects included in the study

IFA ELISA

Subject, disease No.of No. of CSF No. ofserum No. of No. of CSF No. ofserum

patients samples samples

painssamples

samples

patients tested tested patients tested tested

Patient (CM) 45 90 111 45 90 111

Control

TBE 3 3 3 47 10 59

Meningitis of various etiology 21 21 21 33 33 33

Nodemonstrable CNS disease 8 8 8 10 10 10

Multiplesclerosis 0 0 0 16 16 16

Syphilis 7 7 7 25 12 25

Leptospirosis 0 0 0 4 0 5

Infectiousmononucleosis 0 0 0 20 0 20

None (healthy individuals) 63 0 63 120 0 120

serum samples from some patients without demonstrable

CNS disease.

Inother control groups, we analyzed samples of CSFor

serumorbothfrom healthy individualsaged 0 to 90 years or

from patients with multiple sclerosis, infectious

mononu-cleosis, syphilis, or leptospirosis. All controls (except

pa-tients with syphilis) having positive titers to I. ricinus

spirochete in CSF or serum or both were negative by

syphilis serological tests (Wasserman, Venereal Disease

Research Laboratory, fluorescent treponemal

antibody-ab-sorption test).

Serological methods. (i) IFA. The antigen used was a

preparation of spirochetes from SwedishI. ricinus(strainG

152) and wasgrown in modified Kelly medium (14),

centri-fuged at 10,000 x g for 30 min, washed three times in

phosphate-bufferedsaline(PBS),andfixedtoglassslides by

methanolfixation. Twofold dilutions ofserum orCSFwere

made from initial dilutions of1:20forserumand 1:5for CSF.

The slides were covered by these dilutions for 30 min,

washed, and then stained with fluoresceinatedgoat

anti-hu-manimmunoglobulin (polyvalent conjugate; National

Bacte-riological Laboratory, Stockholm,Sweden) for 30min.After

repeated washing, the slides were read in a fluorescence

microscope.Theendpointwasdefinedasthehighest titerat

whichall organisms still showed abright fluorescence. The

same high-titer and low-titer control serum samples were

included in each test.

(ii)ELISA. Theantigenpreparationwasperformed

essen-tially by the methodof Craftetal. (7).

Spirochetes

isolated

from Swedish I. ricinus (strain G 152) were grown in

modifiedKelly medium for4to7days.Thespirocheteswere

thenpelletedbycentrifugation at 10,000x gfor30 min and

washed four timesinPBS with5mM

MgCl2.

The finalpellet

wassuspended in PBS withoutMgCl2and thensonicatedon

icefour times for 30seach time. Afterafinal

centrifugation

at 10,000 x g for 30 min, the supernatant was used as

antigen. Theproteincontentwasdeterminedbythemethod

ofLowry (10).Thecoating concentrationwasdetermined

by

testing a serum with a high concentration of

spirochetal

antibodies against serial dilutions of

antigen.

The

optimal

coating concentrationwasfoundtobe 5,gof

protein

per ml

forimmunoglobulinG(IgG)and 10

,ug/ml

forIgM.

Antigens

were dilutedin PBS (pH7.2).

Microplates (129 B; Dynatech

Laboratories,

Plochingen,

WestGermany)werecoatedovernightat roomtemperature.

Afterrepeated washings, 100,ul ofserum diluted1:1,000or

100,ul of CSF diluted 1:500 in PBS with

0.05%

Tween20was

added, and the plates were incubated for 1 h at room

temperature for IgG and 2 h at 37°C for IgM. Alkaline

phosphatase-conjugated swine anti-human IgG(100 ,ul;Orion

Diagnostica, Helsinki), diluted 1:250, or IgM, diluted 1:200 in PBS with 0.05% Tween 20, was added to each well, and the plates were incubated overnight at room temperature.

p-Nitrophenylphosphate (Sigma Chemical Co., St. Louis,

Mo.), 1 mg/ml in 1 M diethanolamine buffer (pH 9.8)

containing 5 mM MgCl2, was added to each well.

A405

was determined after 25 to 65 min in an ELISA reader (Flow

Laboratories, Irvine, Scotland). All washings weredoneat

least three times with 0.9% NaCl containing 0.05% Tween 20.Positive andnegative controls were included in each test, and the time for substrate incubation was adjusted to these controls to eliminate day-to-dayvariations.

The ELISA titer was defined as the

A405

valuemultiplied by the serum or CSF dilution factor. Each sample wastested

in duplicate, and the meanvalue was calculated. Ifthe two

values differed more than 10% from the mean, the sample

wasretested. If

A405

wasmore than 1.5 or less than 0.1, the

sample was retested in a higher or lower dilution. Each

samplewastested in twouncoated wells(background),and

if the

A405

value in these uncoated wells was higher than

approximately25% of the value for coatedwells,thesample

was retested. All specimens from each patient were tested

simultaneously. Allserum specimens havingantibodytiters

above the 95percentileforhealthyindividuals were

consid-eredsignificantly elevated, i.e., positive.Anincrease in titer

offourfold or greater in IFA or twofold or greater in ELISA

was considered significant. A CSF/serum titer ratio was

expressed as the ratio of the ELISA titer in CSF to the

ELISA titer in serum. Albumin and IgG levels in CSF and serumsamples were measuredby anephelometric method, and aspirochetal CSFtiterindexwascalculatedasthe ratio of(ELISA titer in CSF/ELISA titer inserum)to(albuminin CSF/albuminin serum) (17).

RESULTS

IFA. Spirochetal antibody levels measured by IFA in

serum and CSF samples are shown in

Fig.

1. The 95

percentile serum level inhealthy individuals was 320. With

this limit, 23 of 45

(51%)

CM

patients

were seropositive

when thesample with thehighesttiterwas considered.

CSF samples with detectable

antibodies, i.e.,

a titer of

.1:5,

were considered

positive.

Positive

antibody

titers in

CSFwerefound in 38of 45

(84%)

CM

patients

and innone

of32

meningitis

control

patients.

ELISA. (i) Serum antibodies.

Spirochetal IgG

and IgM

antibodytiters in serum measured

by

ELISA are shown in

on April 11, 2020 by guest

http://jcm.asm.org/

5120 2560 1280 640

160 80 40

20

<20

0@ *0

*000*

,_....

0@@0

0* * *0@ @

Chronic meningitis n=45

Serum

*0

0

Syphilis

n=7

@-0

006

.*

.50..

see

0

0O0

0000

0@@O0

S....0

es....o

Meningitis Healthy controls individuals

n=32 n=63

0

0@

*00

000

S..

00000 000@@O

0@0

0O0

S.. *-..

Chronic meningitis n=45

CSF

S * 0 0*

S

Syphilis

n=7

A640

320 -160 >

80

0

40 _.

20 C

10 5 <5

Meningitis controls n=32

FIG. 1. Antibody levels measuredbyIFA in patientswithCM and control patients. Each dotrepresentsthe highesttiter noted inone

individual. Horizontal barsmark the median titer in eachgroup.The 95percentileof titersinserumforhealthyindividuals usedastheupper limitof normal valueswas320. InCSF,titers of .5wereconsideredpositive.

Fig. 2. The 95 percentile titer level for healthy individuals

was450 for IgG and 580 forIgM. With these limits,11of 120 (9%)healthy controlswere seropositive forIgGorIgM with one patient positive for both IgG and IgM. No significant differenceswerefound indifferent age groups.

The 95 percentile of titer values in a healthy population was set as the upper limit for normal values. Of 45 CM patients, 28 (62%) were IgG positive, 10 (22%) were IgM positive, and 8 (18%)werebothIgGandIgMpositive. Thus, 30 of 45patients(67%)werepositive for IgGorIgMorboth (Fig. 2; see Fig. 4). The median titer levels were not

significantly different between healthy individuals and

pa-tients with TBEormeningitis of variousetiology.Noneof 16 patients with multiple sclerosiswasseropositive, whereas 10 of 20 patients with mononucleosis were IgM seropositive

and 2of 20werebothIgG and IgM positive. Among patients

withotherspirochetal diseases, 19 of 25 (76%) patients with syphilis and 1 of 4 patients with leptospirosis had signifi-cantly elevatedantibody titers againstI. ricinus spirochete. (ii)CSF antibodies. The spirochetal IgG andIgM antibody

levels in CSF measured by ELISA are shown in Fig. 3. A titer levelof 10wasset astheupperlimit for normal values for both IgG and IgM. With these limits, elevation of IgG antibody titers in CSF was found in 39 of 45 CM patients (87%), elevation ofIgM antibody titerswasfound in 29 of 45 (64%), and 41 of 45 patients (91%) showed elevation of either IgG or IgM antibodies orboth (Fig. 3 and 4). Significantly

elevated antibody titers in CSF were found in 29 of 30 seropositive CM patients (datanotshown infigures). Among

themeningitis controls, 3 of 53 (6%)werepositive for IgGor IgM or both. All three patients were also seropositive and showed high CSF/serum albumin ratios, indicating a high degree of damage to the blood-brain barrier. Two of these

threepatients suffered from tuberculousmeningitis, andone patient suffered from varicella-zoster meningoencephalitis. None of 16 patients with multiple sclerosis had elevated antibody titers inCSF, whereas 3 of 12 patients with syphilis wereIgG positive.

CSF/serum ELISA titer ratio and titer index for CSF. An IgG titer in CSF/IgG titer inserumratioof 0.04 andan IgM

titer in CSF/IgM titerinserum ratioof 0.02werechosenas

theupperlimits for normal values. Thecorresponding values for titer indexes for CSFwere 2.0 forIgGand 1.0for IgM. With theselimits, CSF/serum titer ratioswerepositive for 36

of 45(80%) CM patients for IgG, 32 of45(71%) forIgM,and

39 of 45 (87%) for either IgG or IgM or both. The

corre-sponding results for the spirochetal titer index for CSFwere asfollows:of45CMpatients, 39 (87%)werepositive forIgG

antibodies, 32 (71%) were positive for IgM, and 41 (91%) were positive for IgGorIgM orboth (Fig. 4).

Sensitivity and specificity of the differentassays.Theresults

of the differentantibodyassaysand indexcalculationsinCM patients (sensitivity) and meningitis control patients (speci-ficity) areshown inFig. 4. Measurement ofCSF antibodies was moresensitive and specific thanmeasurementofserum

antibodies in both IFA and ELISA. The highest diagnostic sensitivity with 41 of45positive CM patients wasachieved by measuring IgG and IgMantibodies in CSFby ELISA or

by calculating the IgG and IgM spirochetal ELISA titer index for CSF. Three CM patients were negative in all antibody assays and index calculations. One further CM

patientwasseropositive by ELISA for bothIgG and IgM but negative in all other assays and index calculations. Spiro-chetal etiology of disease could thus be found in 41 of 45 (91%) CM patients. The corresponding specificity found for

measurement of CSF antibodies by ELISA was 94%, and

the specificity found for the titer index for CSFwas 98%.

Onlyonemeningitis control patientwaspositive by the titer index calculationfor CSF. Thispatient suffered from tuber-culous meningitis and had a slightly elevated IgM index of 1.03.

Influenceof duration ofdisease andantibiotictreatmenton

titer levels.Levels of ELISA titertospirochete in relationto

duration of disease and antibiotic treatment are shown in Fig. 5. Of 39pretreatment samples, 21 (54%)were seroposi-tive for IgG, 8 (21%) were positive for IgM, and 23 (59%) I..

E w 0.

a)

on April 11, 2020 by guest

http://jcm.asm.org/

SERUM

.0

-.8

4-._5

C)

CO

0)

cn

J

m r-En

Ct

i

1-_.

8

0

m

9

x

Chronic Syphilis Meningitis Healthy mneningitis controls individuals

n=45 n-25 n=90 n=120

Chronic Syphilis Meningitis Healthy meningitis controls individuals n=45 n=25 n=90 n-120

FIG. 2. Antibody levels of IgG and IgM inserummeasuredby ELISA in patients with CM and in control patients. Each dot represents thehighest titer noted in one individual. Horizontal bars mark the median titer in each group. The 95 percentile of titers in serum for healthy individuals used as the upper limit of normal values was 450 for IgG and 580 for IgM.

were positive for IgG or IgM or both. With increasing

duration ofdisease beforethestartoftreatmentwith

antibi-otics, thepercentage of positive CM patients and their IgG

titers increased and theirIgMtiters decreased. After

treat-ment, the titers in CSF and serum usually decreased, with

theexceptionof the IgG titers inserum.Thesetitersshowed

a slight increasing trend during the first5 weeks

posttreat-ment,especiallyinpatients withashort duration ofdisease pretreatment. AmongCM patients with paired serum

sam-ples,asignificantrise ofELISA titerswasfound in only 5 of

35 (14%). In allfivecases, the risewasin the IgGantibody

class. Both of the patients not treated with antibiotics

showed a significant rise in titers. All 32 CM patients with

both pre- and posttreatment CSF samples showed higher

CSF >100.

80.

L-C.)

C

0

-.iE_

0

U-c) 3

cn

cr) 0

0)

w6

40-

20-

10-

s5-0

-0 0 0 0*

0@

00 0

00 S

00 0 000

Chronic meningitis n-45

IgG

0

0

0@

0O

*

0 S.*

Syphilis

n=12

0

0

U00-0

Meningitis controls n=53

Multiple sclesis n-16

*0S

0

100

0*

0

*-*.

IgM

0

0

0 0

00 000000 *00 000 0*000

Chronic meningitis n-45

Syphilis n-12

Meningitis controls n.53

80 m

o

CE

40 0 -n

_.

R

20

-en

c 10 o

C)

EO

§

Multiple scleris n-16

FIG. 3. CSFantibody levels ofIgGandIgMmeasuredby ELISA inpatientswithCMand in controlpatients. Each dotrepresents the highest titer noted inoneindividual. Horizontal barsmarkthemedian titer ineachgroup.Atiterlevel of10wasusedastheupper limitof normalvalues for bothIgGandIgM.

*-

on April 11, 2020 by guest

http://jcm.asm.org/

IFA

U)

cn

0. U1)

_

Co .t_

._E

a)

E

.E-2

:_

0

0

co 0

0.

*E

a)

. o)

a-e° 100

50

20

0c

4 67

Z

0

2

11

lL

Uc

F877

ELISA

U)

a)

0 2

0)

Cm

.-C

0

n

0

0)

o

_O

O)

0

._-.E

Cn')_cn

_D

X4~

0 C .,...

[87T

0

-O

0

0) 0

a

0)

E.

0 0

O)

a)U°

4.Vo v ._

87

7

0) 0)

.i1

.-C n -0

0-C!,

0

z

FIG. 4. Thesensitivity (columns above horizontalbar) andspecificity(columns below horizontalbar)of the differentdiagnosticassaysand calculations in samples from45patients with CM and53meningitis controlpatients(ELISA)or32meningitiscontrolpatients(IFA).IgGor IgM or bothindicatespatients positive for IgGorIgMorboth.

antibody titers in their pretreatment samples than in their

posttreatment samples.

By IFA (data not shown), 21 of 39 (54%) pretreatment

samplesfrom CM patientswere seropositive. With

increas-ing duration of disease before the start of antibiotic

treat-ment, the percentages of CSF- and serum-positive CM

patients increased. Thekinetics ofthetiterdecrease during

the posttreatment period followed essentially the pattern

described for ELISA IgG antibodies. Only 1 of 35 CM

patients with paired serum

samples

showed a

significant

increase in serum titer. None of32 CMpatients with both

pre- and posttreatment CSF samples had higher titers in

their posttreatment samples compared with their

pretreat-mentsamples.

DISCUSSION

Aprobable spirochetal etiology couldbedemonstratedby

ELISA in almost all

patients

with CM examined. The

PRETREATMENT

CSF

II

17k~

igG 90

92ri

IgM

76

67

wvksafterorset

CSFSenau

titer titer

20011000

100 500

POSTTREATMENT

Serum CSF

IgG

igJ

IgO

lgm

80

6 7

4t

I1~

HI'&

19%

IFk

77

25

I

76

44

o,<,

-0089

TP

O T?L.

weeks after treatment

FIG. 5. Kineticsof theantibodyresponse,measuredbyELISA,inrelationto onsetof disease and antibiotictreatment.Columnsrepresent the percentageofpositive patients; 0,mediantiter levels.Seventeen,12, and 10 paired CSF andserumpretreatmentsamplesfrom39patients with 3to5, 6to10, and 11to43weeksduration of disease, respectively, weretested,as were48 CSF and 67serumposttreatmentsamples from 36patients.

%positiveSem Serum

patients

titer

p

50 500

F~

CSF positiwe

titr patienX

200

10050

%13 0

i

T-~

1001000

on April 11, 2020 by guest

http://jcm.asm.org/

sensitivity of the different assaysmight however be

under-estimated, since sensitivity had to be evaluated against a clinical diagnosis of the disease. The possibility of nonspiro-chetal etiology in some patients with CM cantherefore not be completely excluded. With the upper limit of normal valuesestablished inhealthycontrols, the assays also showed a high diagnostic specificity in patients with meningitis of nonspirochetal etiology. However, measurement of CSF antibodies was necessary for diagnosis in approximately 25% of the CM patients. This approach was especially important in patients with short duration ofdisease, since these patients oftenwere seronegative. Generally, measure-ment ofCSFantibodies was a more sensitive method than measurement of serum antibodies byeither IFA or ELISA.

In addition, measurement of CSF antibodies was a more

specific method than measurementofserumantibodies.

The spirochetal antibody response seemed to be rather

slow, andhighantibodylevelsinserum were usually reached between 5 and 10 weeks after the onset of symptoms. Treatment with high-doseintravenous antibiotics seemed to

influence the antibody response in both serum and CSF.

Thismight be due to aninterruption of the antigenic stimu-lation of the immune system resulting in stationary or

decreasingtiters. In patientsreceiving antibiotictreatment,a

significant rise ofserumantibody titers was seldom found. A significant rise oftiters was found in 2 of 2untreatedpatients compared with 3 of 33 patients receiving antibiotics.

The results in this study also indicate that theperformance

ofIgM serologyadds little to the total diagnostic sensitivity

of the ELISA test. However, it is possible that with a tendency towards earlier admission and treatment of the

patients, the IgM serology could become more important.

Serum antibodies to B. burgdorferi in Lyme disease and

related disordershave been analyzed byBurgdorfer (6) and

subsequently by Steere (14) in the United States and by

Ackermann in the Federal Republic of Germany (1), using

IFA. Theyallfound about 90% sensitivity and almost 100% specificity in the neurological stage of the disease. Recently

Russellet al.(11) and Craft et al. (7) have comparedthe IFA

withELISA. In both studies, ELISA was considered

some-what more sensitive and specific than IFA. Thesensitivityof ELISA in serum was >90% in theneurological stage, which is higher than the sensitivity found in this study, possibly

reflectingdifferencesintheselectionof patients andcontrols

and also possibly reflecting differences in the nature of the

diseasesseen in the United States and in Europe.

The high antibody levels in CSF indicated intrathecal

productionofspirochetal antibodiesin theCNS.This

obser-vation agrees well with our previous finding of elevated

synthesis of IgG in the CNS of ourpatients with CM (13).

Also, by imprint immunofixation ofelectrofocused and

se-quentially sampled CSF and serum pairs from one patient

with CM, intrathecal production ofoligoclonal specific IgG

antibodies toB. burgdorferihas been demonstrated (16a).

The demonstration of intrathecal synthesis of specific

antibodies has been utilized not only in chronic neurologic

disorders, especially multiple sclerosis, but also in acute

CNSinfections, especially herpes simplex encephalitis (12,

18, 19). The importanceofconsidering the degreeof

blood-CSF barrier leakage, measured by reference antibodies or

albumin or both, has been pointed out (13).

Against this background we found it relevant to focus

interest on aspirochete-CSF titerindex. Such an index has

someadvantages. A serumantibody titer to spirochetes per

se does not need to be related to the etiology of the actual

disease. A significantly high spirochete-CSF titer index,

however, supports the spirochetal etiology of the CNS

disease. By calculation ofa specific CSF titer index, the

degree

ofblood-CSF barrierdamage maybeconsidered.

An elevated spirochete-CSF titer index was found in a

total of91% of45CM

patients

and alsoin 1of4 patientswith

tuberculous

meningitis.

Twoofthefourpatients with

tuber-culous

meningitis

also had positive titers to our

spirochete

strain in their serum samples. This finding might indicate

some

degree

ofantigenic cross-reactivity between this

spi-rochete strainandMycobacterium sp.

A

cross-reactivity

between serafrom patientswith

syph-ilis and patients with CM was found in this study. This

finding is in agreement with other studies showing

cross-reactivity

between sera from patients with Lyme

disease,

syphilis,

and

relapsing

fever

(7,

11, 14). This

cross-reactivity

iseasily managed in routine diagnostic serological workby

testing

all

specimens

positive by specific spirochetal

serol-ogy

by

routine

syphilis serology

as well. Although

patients

with spirochetal disease were negative by syphilis serology

tests, weakly positive fluorescent treponemal

antibody-absorption

test results may be observed in patients with

Lyme diseaseand related disorders.

In Europe,diagnostic separationofspirochetalmeningitis

from viral TBEis ofspecial interest becauseof thecommon

vector. Furthermore, it may bedifficult to separate the two

diseases on clinical grounds. In this study there was no

cross-reactivity between antibodiesto ourspirochetalstrain

and to TBE viral antigen. We did not find any definite

evidence ofdualinfection, but unfortunately CSFspecimens

were notavailablefrom the patientswith TBE who alsohad

high spirochetal titers inserum.

Inconclusion, the present study shows that theserological

diagnosis of spirochetal orLyme meningitis becomes more

sensitive and specific if CSF serology is performed,

espe-cially in patients withashort duration ofdisease. Themost

specific method is to calculate a spirochetal titer index for

CSFthat takes thedegreeofblood-CSFbarrier damage into

consideration. Antibiotic treatment seems to interrupt the

increase of spirochetal antibody levels and in most cases

eliminates the possibility of serological diagnosis by

deter-mination ofsignificant titer increase.

ACKNOWLEDGMENTS

We thank Arne Svedmyr for supplying specimens from patients with TBE, EivorNordstrom, GerdLundstrom,andMonica Helges-son for technical assistance, Brita Hedberg and RosaPersson for preparingthe manuscript, andEivorBoninfor drawing thefigures.

This workwas supported bygrantsfrom KarolinskaInstitute. LITERATURE CITED

1. Ackermann, R., J. Kabatzki, H. P. Boisten, A. C. Steere,R. L. Grodzicki, S. Hartung, and U. Runne. 1984. Spirochaten-Atiologie der Erythema chronicum migrans Krankheit. Dtsch. Med. Wochenschr. 109:92-97.

2. Afzelius, A. 1910. Report to Verhandlungen der Dermatolo-gische Gesellschaft zu Stockholm on October 28, 1909. Arch. Dermatol. 101:404.

3. Bannwarth, A. 1941. Chronische lymphocytare Meningitis, entzundlichePolyneuritis und "Rheumatismus." Arch. Psychi-atr. Nervenkr. 113:284-376.

4. Bannwarth, A. 1944. ZurKlinik undpathogenese "der chronis-chen lymphocytaren Meningitis." Arch. Psychiatr. Nervenkr. 117:161-185.

5. Benach, J. L., E. M. Bosler, J. P. Hanrahan, J.L. Coleman, G. S. Habicht, T. F. Bast, D. J. Cameron, J. L. Ziegler, A. G. Barbour, W. Burgdorfer, R. Edelman,and R. A. Kaslow. 1983. Spirochetes isolated fromthe bloodoftwo patients withLyme disease. N. Engl. J. Med. 380:740-742.

on April 11, 2020 by guest

http://jcm.asm.org/

6. Burgdorfer, W., A. G.Barbour, S.F. Hayes, J. L.Benach,E. Grunwaldt,andJ. P. Davis. 1982. Lyme disease-a tick-borne spirochetosis? Science 216:1317-1319.

7. Craft, J.E., R. L. Grodzicki, and A. C. Steere. 1983. Antibody response in Lyme disease-evaluation of diagnostic tests. J. Infect. Dis. 149:784-795.

8. Garin, C., and C. Bujadoux. 1922. Paralysie parles tiques. J. Med. Lyon 77:765-767.

9. Johnson, R. C., G. P. Schmid, F. W. Hyde,A. G.Steigerwalt, and D.J.Brenner. 1984.Borreliaburgdorferi sp. nov.: etiologic agentof Lyme Disease. Int.J. Syst. Bacteriol. 34:496-497. 10. Lowry,0.H., N.J. Rosebrough,A. L.Farr, and R. J. Randall.

1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275.

11. Russell,H., J.S.Sampson,G. P.Schmid,H.W.Wilkinson,and B. Plikaytis. 1984. Enzyme-linked immunosorbent assay and indirect immunofluorescenceassayforLyme disease. J.Infect. Dis. 149:465-470.

12. Skoldenberg,B., K.Kalimo, A.Caristrom,M.Forsgren, and P. Halonen. 1981. Herpes simplex encephalitis-a serological fol-low-upstudy. Synthesis of herpes simplex virus immunoglobu-lin M, A and G antibodies and development of oligoclonal immunoglobulin G in the centralnervous system. Acta Neurol. Scand. 63:273-285.

13. Skoldenberg, B., G. Stiernstedt, A. Garde, G. Kolmodin, A. Carlstrom,andC. E. Nord.1983.Chronicmeningitis caused by apenicillin-sensitive microorganism?Lancetii:75-78.

14. Steere, A. C., R. L. Grodzicki, A. N. Kornblatt, J. E. Craft, A. G.Barbour,W.Burgdorfer, G. P.Schmid, E.Johnson, and S.E. Malawista. 1983. The spirochetal etiology of Lyme dis-ease. N.Engl.J. Med. 308:733-740.

15. Steere,A. C., S. E. Malawista, J. A. Hardin, S. Ruddy, P. W. Askenase, and W. A. Andiman. 1977. Erythema chronicum migrans and Lyme arthritis: the enlarging clinical spectrum. Ann.Intern. Med. 86:685-698.

16. Steere, A. C., S. E. Malawista, D. R. Snydman, R. E. Shope, W. A. Andiman, M. R. Ross, and F. M. Steele. 1977. Lyme arthritis: anepidemicofoligoarticulararthritis in children and adults in three Connecticut communities. Arthritis Rheum. 20:7-17.

16a.Stiernstedt, G. T., B. R. Skoldenberg, B. Vandvik, B. Heder-stedt, A.Garde, G.Kolmodin,H.Jorback,and B.Svenungsson. 1984.Chronicmeningitis and Lyme disease in Sweden. YaleJ. Biol.Med. 57:491-497.

17. Tibbling, G., H. Link, and S. Ohman. 1977. Principles of albumin andIgG analysis in neurological disorders.I. Establish-ment of reference values. Scand. J. Clin. Lab. Invest. 37:385-390.

18. Vandvik, B., R. E. Nielsen, F. Vartdal, and E. Norrby. 1982. Mumpsmeningitis: specific and non-specific antibody responses in the centralnervoussystem. ActaNeurol.Scand. 65:468-487. 19. Vartdal, F., B. Vandvik, and E. Norrby. 1980. Viral and bacterialantibodyresponsesinmultiple sclerosis. Ann.Neurol. 8:248-252.

on April 11, 2020 by guest

http://jcm.asm.org/