www.elsevier.com/locate/jns

Journal of the Neurological Sciences 217 (2004) 181–188

Intrathecal production of Chlamydia pneumoniae-specific high-affinity

antibodies is significantly associated to a subset of multiple

sclerosis patients with progressive forms

Enrico Fainardia,b,*, Massimiliano Castellazzia, Ilaria Casettaa, Rosario Cultrerac,Luca Vaghia, Enrico Granieria, Carlo Continic

aMultiple Sclerosis Center, Department of Neurology, University of Ferrara, Arcispedale S. Anna, Corso della Giovecca 203, Ferrara I-44100, ItalybSection of Neuroradiology, Department of Neurosciences, Azienda Ospedaliera-Universitaria, Arcispedale S. Anna, Corso della Giovecca 203,

Ferrara I-44100, ItalycSection of Infectious Diseases, Department of Clinical and Experimental Medicine, University of Ferrara, via Fossato di Mortara, 23, Ferrara I-44100, Italy

Received 15 July 2003; received in revised form 19 September 2003; accepted 24 September 2003

Abstract

The purpose of this study was to provide further insight into the effective relevance of the association between Chlamydia pneumoniae

and MS. We evaluated by ELISA technique cerebrospinal fluid (CSF) and serum levels of anti-C. pneumoniae IgG in 46 relapsing–

remitting (RR), 14 secondary progressive (SP) and 11 primary progressive (PP) MS patients grouped according to clinical and Magnetic

Resonance Imaging (MRI) evidence of disease activity. Fifty-one patients with other inflammatory neurological disorders (OIND) and 52

with non-inflammatory neurological disorders (NIND) were used as controls. A C. pneumoniae-specific intrathecal IgG synthesis as detected

by the relative specific index was present in a small proportion of MS (17%), OIND (22%) and NIND (2%) patients and was significantly

more frequent in MS and in OIND than in NIND ( p < 0.001) and in SP and PP MS than in RR MS patients ( p < 0.02). Among the patients

with C. pneumoniae-specific intratecally produced antibodies, CSF high-affinity anti-C. pneumoniae IgG were found in the majority of SP

or PP MS, occasionally in OIND, but not in RR MS and NIND patients. These findings confirm that the presence of a humoral immune

response to C. pneumoniae within the central nervous system (CNS) is not selectively restricted to MS, but is shared by several

inflammatory neurological conditions. In addition, our results suggest that an intrathecal production of C. pneumoniae-specific high-affinity

IgG can occur in a subset of patients with MS progressive forms in which a C. pneumoniae brain chronic persistent infection may play an

important pathogenetic role.

D 2003 Elsevier B.V. All rights reserved.

Keywords: Chlamydia pneumoniae; Multiple sclerosis; Cerebrospinal fluid; Intrathecal synthesis; Antibody affinity

1. Introduction Recently, a potential role as a causative agent of MS has

Multiple sclerosis (MS) is commonly considered as an

autoimmune chronic inflammatory demyelinating disease of

the central nervous system (CNS) [1]. The exposure to

environmental factors, such as a virus, in combination with

genetic predisposition could be implicated in MS pathogen-

esis [2]. However, direct evidence for an infectious etiology

in MS is still lacking [1,2].

0022-510X/$ - see front matter D 2003 Elsevier B.V. All rights reserved.

doi:10.1016/j.jns.2003.09.012

* Corresponding author. Department of Neurology, Section of

Neurology, University of Ferrara, Corso della Giovecca 203, Ferrara I-

44100, Italy. Tel.: +39-532-236304; fax: +39-532-205525.

E-mail address: henryfai@tin.it (E. Fainardi).

been suggested for Chlamydia pneumoniae, a Gram-nega-

tive, obligate intracellular bacterium that infects and repli-

cates within monocytes, macrophages and endothelial cells,

can induce a chronic infection, and presents an high

seroprevalence in adults [3]. However, the initial high rates

reported for molecular and culture demonstration of C.

pneumoniae in cerebrospinal fluid (CSF) of MS patients

[4] were not reproduced in the subsequent studies in which

CSF evidence of C. pneumoniae in MS was either less

frequent than previously described [5–9] or absent [10–

13]. In addition, whereas C. pneumoniae was not identified

in MS brain lesions [14,15], a C. pneumoniae-specific

epitope was able to induce an experimental autoimmune

E. Fainardi et al. / Journal of the Neurological Sciences 217 (2004) 181–188182

encephalomyelitis, the animal model of MS, in rats [16].

Contradictory findings have also been reported when the C.

pneumoniae-specific intrathecal humoral immune response

was evaluated. A quantitative production of anti-C. pneu-

moniae antibodies within the CNS was found in approxi-

mately 20–30% of MS patients and, to a lesser extent, in

patients with other neurological disorders [9,13,17]. CSF-

restricted oligoclonal IgG bands specific for C. pneumoniae

were detected in over 80% [18] or in none of MS patients

[13]. The assessment of the C. pneumoniae-specific anti-

body production in the brain is crucial to understand the

effective relevance of this pathogen in MS because it is

well known that, while in MS a polyspecific intrathecal

immune response composed by anti-viral low-affinity anti-

bodies occurs, in infectious diseases intrathecally synthe-

sized high-affinity antibodies, of which at least 30%

specifically directed against the causative agents, predom-

inate [19–22].

Thus, in order to clarify the significance of C. pneumo-

niae infection in MS, in the present study we addressed the

affinity distributions of intrathecally produced anti-C. pneu-

moniae antibodies in MS patients and controls.

2. Patients and methods

2.1. Patient characteristics

Seventy-one consecutive patients (51 women and 20

men; mean age = 40.3F 11.4) satisfying the criteria for

clinically or laboratory-supported definite MS according

to the classification of Poser et al. [23] and followed by

the MS Centre of the Department of Neurology, University

of Ferrara during the period from June 1997 to December

2001 were retrospectively selected for the study. Of these,

46 were relapsing–remitting (RR), 14 were secondary

progressive (SP) and 11 were primary progressive (PP)

in agreement with the criteria of Lublin and Reingold [24].

Relapse was defined as the onset of new or recurrent

symptoms or signs or the worsening of already present

neurological abnormalities persisting for at least 24 h in

the absence of fever and preceded by at least 1 month of

stable or improved neurological state [23]. Evidence of a

relapse at admission was considered as clinical disease

activity. Accordingly, at the time of enrolment in the study

39 (32 RR and 7 SP) MS were clinically active, while 21

(14 RR and 7 SP) MS were clinically stable. Given that PP

MS patients, by definition [24], were not experiencing

relapses these subjects were excluded by these two sub-

groups. Disease severity was scored in all MS patients at

the time of sample collection using Kurtzke’s Expanded

Disability Status Scale (EDSS) [25] (mean at entry =

3.2F 1.3; range from 0 to 6). The duration of the disease

was expressed in months (mean at entry = 73.1F 84.3;

range from 2 to 460). Patients treated with immunosup-

pressive (e.g. azathioprine or methylprednisolone) or im-

munomodulatory (e.g. interferon-beta or glatiramer acetate)

drugs within 6 months before entering the study were

excluded.

Fifty-one patients with other inflammatory neurological

disorders (OIND) and 52 subjects with other non-inflam-

matory neurological disorders (NIND) were selected as

neurological controls. Among the OIND patients (24 wom-

en and 27 men, mean age = 50.96F 15.99), 14 had chronic

inflammatory demyelinating polyneuropathy, 8 had viral

encephalomyelitis, 7 had acute inflammatory demyelinating

polyneuropathy, 5 had bacterial meningitis, 5 had optic

neuritis, 4 had HIV encephalopathy, 4 had neurolupus, 3

had neuroSjogren and 1 had neuroBechet. None of the five

patients with optic neuritis had CSF oligoclonal bands or

Magnetic Resonance Imaging (MRI) MS-like lesions sug-

gesting a future development of MS. Among the NIND

patients (25 women and 27 men, mean age = 47.96F 17.50)

8 had transient ischemic attack, 7 had epilepsy, 6 had

headache, 5 had cervical spondylosis, 5 had hereditary

ataxia, 5 had vascular dementia, 4 had migraine, 4 had

amyotrophic lateral sclerosis, 4 had compression neuropa-

thy, 2 had parestesias, and 2 had Alzheimer disease. All

OIND and NIND patients were not undergoing treatment

with immunosuppressant drugs, including steroids, at the

time of sampling.

2.2. Handling of CSF and serum samples

CSF samples were obtained by atraumatic lumbar punc-

ture performed with appropriate clinical indications. Both

CSF and serum samples were collected under sterile con-

ditions, coded, frozen and stored in aliquots at � 70 jCuntil the assay.

The study was approved by the Regional Committee for

Medical Ethics in Research. CSF and serum IgG and

albumin levels were measured by immunochemical nephe-

lometry with the Beckman Array Protein System (Beckman

Instruments, Fullerton, CA, USA) according to the proce-

dure of Salden et al. [26] (Salden et al., 1988). Blood-CSF-

barrier (B-CSF-B) dysfunction was determined by CSF/

serum albumin quotient (QAlb) [27].

2.3. MRI procedures

All MS patients underwent brain MRI scans at entry on a

1-Tesla MRI unit (GE Signa Horizon, General Electric

Medical Systems, Milwaukee, WI). Routinely used T1-

weighted axial spin echo images were obtained approxi-

mately 10 min after intravenous injection of 0.1 mmol/kg of

Gd-DTPA in each patient. Lesions showing Gd-enhance-

ment on T1-weighted scans were defined as active. MRI

evidence of disease activity was present in 30 (24 RR, 3 SP

and 3 PP) MS patients, whereas in 41 (22 RR, 11 SP and

8 PP) MS patients it was not detected. All brain MRI scans

were evaluated by one investigator (EF) blinded to clinical

and sample data.

Table 1

Serum and CSF detectable levels of anti-C. pneumoniae IgG in patients

with Multiple Sclerosis (MS), Other Inflammatory Neurological Disorders

(OIND) and Non-Inflammatory Neurological Disorders (NIND)

Serum CSF

Total MS (n= 71) 63/71 (88.7%)* 58/71 (81.7%)y

OIND (n= 51) 42/51 (82.3%)# 44/51 (86.3%)z

NIND (n= 52) 33/52 (63.5%) 36/52 (69.2%)

*Total MS vs. NIND ( p< 0.001).y Total MS vs. NIND ( p< 0.02).# OIND vs. NIND ( p< 0.05).z

E. Fainardi et al. / Journal of the Neurological Sciences 217 (2004) 181–188 183

2.4. Elisa assay for serum and CSF levels of anti-C.

pneumoniae IgG

As recently described [28], CSF and serum concentra-

tions of anti-C. pneumoniae specific IgG were measured by

enzyme-linked immunosorbent assay (ELISA) by using a

commercially available ELISA kit (C. pneumoniae IgG

ELISA, Bioclone Australia, Cat. No. 40CPG0096).

2.5. Calculation of C. pneumoniae-specific antibody index

According to Reiber’s formula [29] production of anti-C.

pneumoniae IgG was determined by Antibody Specific

Index (ASI). ASI means the ratio between CSF and serum

AU values (QSpec) and CSF/serum total IgG levels expressed

as mg/dl (QIgG) in accordance with the formula:

ASI ¼ QSpec=QIgG ð1Þ

with QSpec =AUCSF/AUserum and QIgG = IgGCSF/IgGserum.

In the case of intrathecal synthesis of total IgG, QIgG

appeared more elevated than Reiber’s hyperbolic discrimi-

nation line (QLim) [30]. QLim represented the completely

OIND vs. NIND ( p< 0.01).

Fig. 1. Distribution of Antibody Specific Index (ASI) abnormal values for anti-C

remitting (RR) MS, 14 secondary progressive (SP) MS, 11 primary progressive (P

Inflammatory Neurological Disorders (NIND) patients, with higher rates in Total M

(*p< 0.001), in PP MS than in RR MS (^p< 0.02) and in NIND (jp< 0.01) and

blood-derived CSF total IgG fraction calculated from the

individual QAlb of a single patient. Therefore, owing to the

introduction of this barrier-related correction of the QIgG,

the ratio became between QSpec and QLim in agreement with

the formula [31]:

ASI ¼ QSpec=QLim ð2Þ

with QLim ¼0:93

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiðQAlbÞ2 þ 6� 10�6 � 1:7� 10�3

q

Therefore, to avoid false-negative results Eq. (1) (QIgG>

QLim) was used when no significant intrathecal total IgG

synthesis occurred, while Eq. (2) (QIgG <QLim) was pre-

ferred in case of intense brain-derived total IgG production

in CSF.

C. pneumoniae-specific intrathecal IgG synthesis was

assumed for values of ASI greater than 1.5.

2.6. Determination of affinity distributions of CSF anti-C.

pneumoniae IgG

Affinity distributions of anti-C. pneumoniae IgG were

measured in CSF by performing an ELISA protocol based

on the employment of increasing concentrations of sodium

thioyanate (NaSCN). All the assays were carried out using

plates with 8 rows� 12 columns of the above-mentioned

ELISA kit (C. pneumoniae IgG ELISA, Bioclone Aus-

tralia, Cat. No. 40CPG0096). In agreement with the

previous C. pneumoniae-specific IgG measurements, the

same CSF dilutions were utilized. Briefly, 100 Al of eachCSF sample 1:2 diluted in 0.4% saline were added to all

12 wells of each the eight rows of the precoated ELISA

microplate. After 1 h incubation at 37 jC and three

washing cycles, 12 consecutive dilutions of NaSCN

(range: 0–5 M in 0.4% saline) were placed in each row

and incubated for 10 min at 37 jC. The different concen-

. pneumoniae IgG in 71 Total multiple sclerosis (Total MS), 46 relapsing

P) MS, 51 Other Inflammatory Neurological Disorders (OIND) and 52 Non

S than in NIND (jp< 0.01), in SP MS than in RR MS (^p< 0.02) and NIND

in OIND than in RRMS (#p< 0.05) and in NIND (*p< 0.001).

Table 2

Clinical features and C. pneumoniae-specific CSF affinity findings in

Multiple Sclerosis (MS) patients and controls with intrathecally synthesized

IgG anti-C. pneumoniae as indicated by the presence of antibody specific

index values >1.5

Total MS

(n= 12)

OINDa

(n= 11)

NIND

(n= 1)b

Sex (male/female) 3/9 8/3 0/1

Age, years (meanF SD) 42.17F 11.68 60F 11.87 23

Disease duration,

months (meanF SD)

108.33F 79.01

Disease severity,

EDSS (meanF SD)

4.0F 1.0

Clinical course

Relapsing–

Remitting (n)

3/12 (25%)

Secondary

progressive (n)

5/12 (45.4%)

Primary

progressive (n)

4/12 (33.3%)

Clinical activityc (n) 3/12 (25%)

MRI activityd (n) 4/12 (33.3%)

CSF RA at 5 M

NaSCNe (%)

Range 0–8.1 0–3.4 0

Median 3.15 0 0

CSF affinity ratiof

(meanF SD)

Range 0–198.2 0–77.7 0

Median 31.05 0 0

a OIND=Other Inflammatory Neurological Disorders.b NIND=Non Inflammatory Neurological Disorders.c Clinical activity = presence of relapse at entry.d MRI activity =MRI appearance of gadolium enhancing lesions.e CSF RA at 5 M NaSCN=CSF with Relative Affinity (RA) value at

the highest concentration (5.0 M) of sodium thioyanate expressed in

percentages.f CSF affinity ratio (AR) = a single numerical value which relates the

percentages of high affinity antibodies to the percentages of low affinity

antibodies in CSF containing anti-C. pneumoniae IgG.

E. Fainardi et al. / Journal of the Neurological Sciences 217 (2004) 181–188184

trations of NaSCN we applied into the 12 wells of each

row from the first to the twelfth well were: 0, 0.5, 0.75,

1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, and 5 M, where 0 M

corresponding to 0.4% saline buffer. After incubation,

NaSCN was removed and after three washing cycles wells

were blocked by adding 100 Al/wells of 3% bovine serum

albumin (BSA) for 30 min at 37 jC. After incubation,

BSA was discarded and detection antibody reaction and the

colour development were performed as described in

ELISA assay for serum and CSF levels of anti-C. pneumo-

niae IgG. Appropriate background wells (blank) were

included in each test and their optical density values were

subtracted from all readings. The affinity distribution

histograms for CSF C. pneumoniae-specific IgG were

obtained assuming that, for each CSF sample, the results

observed in the first well, in which NaSCN were absent,

represented the total antigen-specific IgG. In each row, we

divided the difference in OD values between two succes-

sive wells (e.g. well 1 and well 2) by OD values of the

total antigen-specific IgG, and multiplied by 100. In this

way, the variation in OD values between two subsequent

wells was expressed as percentage of OD values of the

total antigen-specific IgG measured in the first well. The

difference in OD values between two consecutive wells

reflected the proportion in which the bounded IgG mole-

cules were reduced by the higher concentration of NaSCN

dispensed in the two wells, and the corresponding percent-

age was considered as the Relative Affinity (RA) of

antibodies. When OD values measured in the well with

higher levels of NaSCN were slightly different compared

to those detected in the adjacent well with lower levels of

NaSCN, the RA was referred as zero. When positive OD

values were still measurable in the twelfth well, we

compared these OD values with those of a simulated

thirteenth well which was considered to be zero. Based

on this principle, we obtained an RA value showing the

proportion of antigen-specific IgG which was not removed

by the highest concentration (5.0 M) of NaSCN, and then

the highest antibody affinity. Thus, the affinity histogram

for each CSF sample was generated by plotting the relative

affinities on the x-axis and the corresponding proportion of

IgG expressed as percentages on y-axis. According to

Luxton et al. [21], CSF samples with RA value more than

2.5% from well number twelve containing the highest

concentration (5.0 M) of NaSCN were considered as

having high-affinity anti-C. pneumoniae IgG, whereas

CSF samples with RA value equal or less than 2.5% at

the highest concentration of 5 M NaSCN were regarded as

showing C. pneumoniae-specific low-affinity IgG. In ad-

dition, to increase the sensitivity in the detection of

antibody binding affinity, the affinity of CSF anti-C.

pneumoniae IgG was also defined by Affinity Ratio

(AR), a single numerical value which relates the percen-

tages of high affinity antibodies, expressed by the summa-

tion of the RA values of wells 11 and 12, to the

percentages of low affinity antibodies, represented by the

summation of the RA values of wells 1 through 4,

according to the formula [21]:

AR ¼

XRA11�12

XRA1�4

þ0:1

8>>>><>>>>:

9>>>>=>>>>;

� 100:

2.7. Statistics

Mann–Whitney U-test was used to compare mean values

among the various groups. The patient group percentages

were compared using the Log-Likelihood Ratio Test (G).

The Spearman rank correlation coefficient test was used to

identify possible relationships among the different variables

in the MS group. A value of p < 0.05 was accepted as

statistically significant.

E. Fainardi et al. / Journal of the Neurological Sciences 217 (2004) 181–188 185

3. Results

3.1. Serum and CSF levels and intrathecal synthesis of anti-

C. pneumoniae IgG in MS patients and controls

Detectable serum and CSF levels of anti-C. pneumoniae

IgG were statistically more frequent in total MS

( p < 0.001 and p < 0.02, respectively) and in OIND

( p < 0.05 and p < 0.01, respectively) than in NIND

patients, without any significant differences between total

MS and OIND (Table 1). In contrast, there were no

significant differences for serum and CSF anti-C. pneumo-

niae IgG mean concentrations among total MS, OIND and

NIND groups also when MS patients were evaluated

according to clinical course, and clinical and MRI activity

(data not shown).

Fig. 2. Panel A describes the distribution CSF high affinity IgG to C. pneumon

sclerosis (Total MS), 3 relapsing remitting (RR) MS, 5 secondary progressive

Neurological Disorders (OIND) with ASI values >1.5. CSF high affinity anti-C. pn

SP (*p< 0.02) and PP (jp< 0.05) MS than in RR MS. Panel B reports CSF mean

relates the percentages of high affinity antibodies to the percentages of low affinity

of MS patients and controls. CSF Affinity Ratio mean levels are higher (*p< 0.05)

represents the 25th–75th quartile divided by the median, the whiskers the range.

A C. pneumoniae-specific intrathecal IgG synthesis as

suggested by the presence of ASI abnormal values was

significantly more represented ( p< 0.001) in total MS (12/

71; 16.9%) and in OIND (11/51; 21.6%) than in NIND (1/

52; 1.9%) (Fig. 1). When MS patients were stratified

according to clinical types, and clinical and MRI activity,

ASI values greater than 1.5, indicating an intrathecal

production of anti-C. pneumoniae IgG were significantly

more frequent in SP and PP ( p < 0.02) than in RR MS, in

SP ( p < 0.001) and in PP MS ( p < 0.01) than in NIND and

in OIND ( p < 0.05) than RR MS, with no statistical differ-

ences between SP and PP MS (Fig. 1). No significant

discrepancies were observed for intrathecal synthesis of

anti-C. pneumoniae IgG between clinically active and

clinically stable MS as well as between MS with and

without MRI appearance of disease activity (data not

iae, as reflected by RA value at 5 M NaSCN>2.5%, in 12 Total multiple

(SP) MS, 4 primary progressive (PP) MS and 11 Other Inflammatory

eumoniae IgG is more frequent in Total MS than in OIND (jp< 0.05), and invalues of CSF Affinity Ratio, considered as a single numerical value which

antibodies in CSF containing anti-C. pneumoniae IgG, in the same subsets

in total MS than in OIND, and in SP and PP MS than in RR MS. The box

E. Fainardi et al. / Journal of the Neurolo186

shown). With regard to OIND and NIND, there were no

significant differences for serum and CSF anti-C. pneumo-

niae IgG detectable and mean concentrations as well as for

C. pneumoniae-specific intrathecal IgG synthesis among the

different clinical entities included in these two groups (data

not shown).

3.2. Affinity distributions of intrathecally synthesized anti-

C. pneumoniae IgG

To test the strength of binding of C. pneumoniae-specific

intrathecally produced antibodies, affinity distributions of

IgG directed against C. pneumoniae were studied in all 24

CSF samples (12 MS, 11 OIND and 1 NIND) with an

intrathecal production of anti-C. pneumoniae IgG (Table 2)

and in 12 CSF samples (9 MS, 2 OIND and 1 NIND)

without evidence of C. pneumoniae-specific intrathecal IgG

synthesis in which levels of anti-C. pneumoniae IgG were

detectable in CSF. We also tested, as controls, 3 CSF

samples (1 MS, 1 OIND and 1 NIND) without measurable

CSF levels of anti-C. pneumoniae IgG.

As illustrated in Fig. 2, panel A, in patients with ASI

values >1.5, CSF high-affinity IgG to C. pneumoniae, as

indicated by RA value at 5 M NaSCN>2.5%, was detected

more frequently in total MS (7/12) than in OIND (2/11) and

in NIND (0/1) patients, with a significant difference

( p < 0.05) between total MS and OIND. The two OIND

patients with CSF high-affinity anti-C. pneumoniae IgG

showed one a bacterial meningitis, and the other an acute

inflammatory demyelinating polyneuropathy. As affinity

measurements were performed in only one NIND patient,

this group was not statistically compared with total MS and

OIND. Among the different MS clinical courses, an RA

value at 5 M NaSCN>2.5% was present in the majority of

SP (4/5) and PP (3/4) MS and in none of RR (0/3) MS

patients. Consequently, CSF high-affinity anti-C. pneumo-

niae IgG were significantly more frequent in SP MS

( p < 0.02) and PP MS ( p < 0.05) than in RR MS. Similar

proportions of CSF high-affinity anti-C. pneumoniae IgG

were found in clinically active (2/3; 66.3%) and in clini-

cally stable (2/5; 40%) MS, in MS patients with Gd

enhancing lesions (2/4; 50%) and in those without MRI

focal activity (5/8; 62.5%). Accordingly, CSF Affinity Ratio

mean values were higher ( p < 0.05) in total MS than in

OIND, and in SP and PP MS than in RR MS (Fig. 2, panel

B), with no significant differences among the remaining

groups evaluated.

All patients with ASI values < 1.5 and with detectable C.

pneumoniae-specific IgG levels in CSF, showed CSF low-

affinity IgG to C. pneumoniae, as suggested by RA value at

5 M NaSCN< 2.5%, and comparable CSF Affinity Ratio

mean values (data not shown).

No antibodies with affinity for C. pneumoniae were

demonstrated in CSF of the three controls without measur-

able CSF levels of anti-C. pneumoniae IgG (data not

shown).

3.3. Correlation between CSF and serum C. pneumoniae-

specific humoral immune response and MS clinical features

In MS patients, CSF and serum levels and intrathecal

synthesis of anti-C. pneumoniae IgG, and affinity distribu-

tions of intrathecally synthesized anti-C. pneumoniae IgG

were not significantly correlated to severity, as expressed by

EDSS, and duration of the disease (data not shown).

gical Sciences 217 (2004) 181–188

4. Discussion

We investigated whether the appearance of an intrathecal

production of anti-C. pneumoniae IgG in MS represents a

manifestation of a C. pneumoniae-specific brain infection.

In our study, an intrathecal release of anti-C. pneumoniae

IgG was found in a small proportion of MS (17%), OIND

(22%) and NIND (2%) patients and, as opposed to previous

reports [10,13,17], it was significantly more prevalent in MS

and OIND than in NIND and in SP and PP MS than in RR

MS. In agreement with prior studies [9,13,17], MS patients

with and without relapse as well as with and without Gd

enhancing lesions produced intrathecally the same amount

of anti-C. pneumoniae IgG. In contrast with Sriram et al. [4]

but according to several other publications [5,9–13,17,32],

no significant differences were detected for CSF and serum

mean concentrations of anti-C. pneumoniae IgG among MS,

OIND and NIND. As previously reported [17], we identified

no statistical correlation between CSF and serum levels and

intrathecal synthesis of anti-C. pneumoniae IgG and severity

or duration of the disease. Discrepancies between our

findings and those obtained by other groups may be due

to the use of different procedures during the assay and to the

existence of distinct study designs.

Overall, these findings do not support a central role for

C. pneumoniae in the pathogenesis of all MS patients and

confirm that a C. pneumoniae-specific intrathecal humoral

immune response can be detected in a subgroup of MS and

OIND patients [9,13,17,32], strengthening the assumption

that the presence of C. pneumoniae within the CNS is not

selectively associated with MS, but may be linked to a large

number of inflammatory neurological conditions [8,9,32]. In

addition, the lack of differences observed in our report for

intrathecal synthesis of anti-C. pneumoniae IgG between

MS and OIND do not corroborate the recent hypothesis that

the clearance of C. pneumoniae from CSF could be closely

reduced in MS [8,9,32].

Of particular relevance was the observation that in MS a

C. pneumoniae-specific intrathecal IgG synthesis was pre-

ponderant in a subset of patients with MS progressive

forms. Although SP and PP MS differ in clinical, genetic,

immunological, pathological and radiological features

[33,34], the strong intrathecal production of anti-C. pneumo-

niae IgG we found in a part of both these clinical types of

MS was not unexpected since SP and PP MS share not only

some immunological patterns [34] but also similar mecha-

E. Fainardi et al. / Journal of the Neurological Sciences 217 (2004) 181–188 187

nisms leading to progression [33]. Furthermore, this result

was in agreement with the frequent culture demonstration of

C. pneumoniae previously described in CSF of progressive

MS [4]. Given that in progressive forms of MS there appears

a chronic immune activation sustained by a continual

antigen presentation within the CNS [35], an intrathecal

synthesis of C. pneumoniae-specific IgG may mirror a

persistent infection which promotes an autoimmune re-

sponse within the CNS [2,35].

In the present report we assessed, for the first time, the

affinity distributions of CSF anti-C. pneumoniae IgG in a

subgroup of patients with C. pneumoniae-specific intrathe-

cally produced antibodies. In the course of an infection, the

process of immune maturation requires a progressive in-

crease in antibody affinity that leads to an intensive pro-

duction of high-affinity antibodies specifically directed

against the causative pathogen [36]. In addition, neurolog-

ical infectious diseases are characterized by the presence of

CSF antibodies with high-affinity against the causative

agent [21]. Our principal finding was that, among the

patients with C. pneumoniae-specific intratecally produced

antibodies, CSF high-affinity anti-C. pneumoniae IgG were

found in 7 out of 12 (58.3%) MS patients, all belonging to

SP or PP MS clinical types, and in 2 out of 11 (18.2%)

OIND subjects. Accordingly, AR mean values were more

consistent in total MS than in OIND as well as in SP and PP

MS than in RR MS. These results raise the possibility that

an intrathecal production of C. pneumoniae-specific high-

affinity antibodies, suggesting a C. pneumoniae brain chron-

ic persistent infection, is present in a subset of patients with

MS progressive forms and, more occasionally, in neuro-

inflammatory diseases which have been related to C. pneu-

moniae infection [3].

A contribution of C. pneumoniae to progressive disabil-

ity in MS, however, is not supported by the lack of a

significant correlation between CSF high-affinity anti-C.

pneumoniae IgG and EDSS. This study does not clarify

whether the presence of intrathecal high-affinity anti-C.

pneumoniae antibodies reflects a continual antigenic stimu-

lation sustained by a specific brain chronic persistent

infection or whether it is due to the reactivation of C.

pneumoniae in macrophages and endothelial cells promoted

by the overactive immune response operating in progressive

forms of MS [35]. A causal association between CSF high-

affinity specific-antibodies and CNS infection has been

demonstrated only in acute and not in chronic encephalitis

[21]. In addition, we did not evaluate ASI and CSF affinity

distributions of specific-IgG for other infectious agents,

such as measles, rubella, varicella zoster, frequently

detected in CSF of MS patients, thus questioning the

specificity of C. pneumoniae antibodies we found in CSF

for progressive MS.

The hypothesis suggested in this study requires more

extensive research and a larger number of patients before

definite conclusions can be drawn on the role of C. pneumo-

niae in MS.

Acknowledgements

The study was supported by Fondazione Cassa di

Risparmio di Ferrara, Italy. We thank Dr. Tiziana Bellini

and Professor Franco Dallocchio of the Biochemistry

Department of University of Ferrara and Dr. Roberto Furlan

of Neuroimmunology Unit, Department of Neurology, San

Raffaele Scientific Institute of Milan for their insightful

discussions and thoughtful suggestions in the development

of ELISA method to measure antibody relative affinity in

CSF.

References

[1] Noseworthy JH, Lucchinetti C, Rodriguez M, Weinshenker BG. Mul-

tiple sclerosis. N Engl J Med 2000;343:938–52.

[2] Meinl E. Concepts of viral pathogenesis of multiple sclerosis. Curr

Opin Neurol 1999;12:303–7.

[3] Yucesan C, Sriram S. Chlamydia pneumoniae infection of the central

nervous system. Curr Opin Neurol 2001;14:355–9.

[4] Sriram S, Stratton CW, Yao SY, Tharp A, Ding L, Bannan JD, et al.

Chlamydia pneumoniae infection of the central nervous system in

multiple sclerosis. Ann Neurol 1999;46:6–14.

[5] Layh-Schmitt G, Bend C, Hildt U, Dong-Si T, Juttler E, Schnitzler P,

et al. Evidence for infection with Chlamydia pneumoniae in a sub-

group of patients with multiple sclerosis. Ann Neurol 2000;47:652–5.

[6] Treib J, Haaß A, Stille W, Maass M, Stephan C, Holzer G, et al.

Multiple sclerosis and Chlamydia pneumoniae. Ann Neurol 2000;

47:408.

[7] Ikejima H, Haranaga S, Takemura H, Kamo T, Takahashi Y, Friedman

H, et al. PCR-based method for isolation and detection of Chlamydia

pneumoniae DNA in cerebrospinal fluid. Clin Diagn Lab Immunol

2001;8:499–502.

[8] Gieffers J, Poh D, Treib J, Dittmann R, Stephan C, Klotz K, et al.

Presence of Chlamydia pneumoniae DNA in the cerebral spinal fluid

is a common phenomenon in a variety of neurological diseases and

not restricted to multiple sclerosis. Ann Neurol 2001;49:585–9.

[9] Hao Q, Miyashita N, Matsui M, Wang HY, Matsushima T, Saida T.

Chlamydia pneumoniae infection associated with enhanced MRI spi-

nal lesions in multiple sclerosis. Mult Scler 2002;8:436–40.

[10] Boman J, Roblin PM, Sundstrom P, Sandstrom M, Hammerschlag

MR. Failure to detect Chlamydia pneumoniae in the central nervous

system of patients with MS. Neurology 2000;54:265.

[11] Pucci E, Taus C, Carthechini E, Morelli M, Giuliani G, Clementi M,

et al. Lack of Chlamydia infection of the central nervous system in

multiple sclerosis. Ann Neurol 2000;48:340–99.

[12] Saiz A, Marcos MA, Graus F, Vidal J, Jimenez de Anta MT. No

evidence of CNS infection with Chlamydia pneumoniae in patients

with multiple sclerosis. J Neurol 2001;248:617–8.

[13] Derfuss T, Gurkov R, Bergh FT, Goebels N, Hartmann M, Barz C, et

al. Intrathecal antibody production against Chlamydia pneumoniae in

multiple sclerosis is part of a polyspecific immune response. Brain

2001;124:1325–35.

[14] Hammerschlag MR, Ke Z, Lu F, Roblin P, Boman J, Kalman B. Is

Chlamydia pneumoniae present in brain lesions of patients with multi-

ple sclerosis? J Clin Microbiol 2000;38:4274–6.

[15] Morre SA, De Groot CJA, Killestein J, Meijer CJLM, Polman CH,

Van der Valk P, et al. Is Chlamydia pneumoniae present in the central

nervous system of multiple sclerosis patients? Ann Neurol 2000;

48:399.

[16] Lenz DC, Lu L, Conant SB, Wolf NA, Gerard HC, Whittum-Hudson

JA, et al. A Chlamydia pneumoniae-specific peptide induces experi-

mental autoimmune encephalomyelitis in rats. J Immunol 2001;

167:1803–8.

E. Fainardi et al. / Journal of the Neurological Sciences 217 (2004) 181–188188

[17] Krametter D, Niederwieser G, Berghold A, Birnbaum G, Strasser-

Fuchs S, Hartung HP, et al. Chlamydia pneumoniae in multiple scle-

rosis: humoral immune responses in serum and cerebrospinal fluid

and correlation with disease activity marker. Mult Scler 2001;7:13–8.

[18] Yao SY, Stratton CW, Mitchell WM, Sriram S. CSF oligoclonal bands

in MS include antibodies against Chlamydiophila antigens. Neurol-

ogy 2001;56:1168–76.

[19] Sindic CJM, Monteyne P, Laterre EC. The intrathecal synthesis of

virus-specific oligoclonal IgG in multiple sclerosis. J Neuroimmunol

1994;54:75–80.

[20] Conrad AJ, Chiang EY, Andeen LE, Avolio C, Walker SM, Baum-

hefner RW, et al. Quantitation of intrathecal measles virus IgG anti-

body synthesis rate: subacute sclerosing panencephalitis and multiple

sclerosis. J Neuroimmunol 1994;54:99–108.

[21] Luxton RW, Zeman A, Holzel H, Harvey P, Wilson J, Kocen R, et al.

Affinity of antigen-specific IgG distinguishes multiple sclerosis from

encephalitis. J Neurol Sci 1995;132:11–9.

[22] Reiber H, Ungefehr S, Jacobi C. The intrathecal, polyspecific and

oligoclonal immune response in multiple sclerosis. Mult Scler 1998;

4:111–7.

[23] Poser CM, Paty DW, Scheimberger L, McDonald WI, Davis FA,

Ebers GC, et al. New diagnostic criteria for MS: guidelines for re-

search protocols. Ann Neurol 1983;13:227–31.

[24] Lublin DF, Reingold SC. Defining the clinical course of multiple

sclerosis: results of an international survey. Neurology 1996;46:

907–11.

[25] Kurtzke JF. Rating neurological impairment in multiple sclerosis: an

expanded disability scale (EDSS). Neurology 1983;33:1444–52.

[26] Salden HJM, Bas BM, Hermas JTH, Janson PCV. Analytical perform-

ance of three commercially available nephelomethers compared quan-

tifying protein in serum and cerebrospinal fluid. Clin Chem 1988;

34:1594–6.

[27] Tibbling G, Link H, Ohman S. Principles of albumin and IgG anal-

yses in neurological disorders: I. Establishment of reference value.

Scand J Clin Lab Invest 1977;37:385–90.

[28] Contini C, Fainardi E, Seraceni S, Granieri E, Castellazzi M, Cultrera

R. Molecular identification and antibody testing of Chlamydiophila

pneumoniae in a subgroup of patients with HIV-associated dementia

complex preliminary results. J Neuroimmunol 2003;136:172–7.

[29] Reiber H, Lange P. Quantification of virus-specific antibodies in cere-

brospinal fluid and serum: sensitive and specific detection of antibody

synthesis in brain. Clin Chem 1991;37:1153–60.

[30] Reiber H, Felgenhauer K. Protein transfer at the blood cerebrospinal

fluid barrier and the quantitation of the humoral immune response

within the central nervous system. Clin Chem Acta 1987;163:319–28.

[31] Reiber H. Evaluation of blood cerebrospinal fluid barrier function and

quantification of the humoral immune response within the central

nervous system. In: Thompson EJ, Trojano M, Livrea P, editors.

Cerebrospinal fluid analysis in multiple sclerosis. Milano: Springer;

1996. p. 51–72.

[32] Sotgiu S, Piana A, Pugliatti M, Sotgiu A, Deiana GA, Sgaramella E,

et al. Chlamydia pneumoniae in the cerebrospinal fluid of patients

with multiple sclerosis and neurological controls. Mult Scler 2001;

7:371–4.

[33] Thompson AJ, Polman CH, Miller DH, Mc Donald WI, Brochet B,

Filippi M, et al. Primary progressive multiple sclerosis. Brain 1997;

120:1085–96.

[34] McDonnell GV, Hawkins SA. Primary progressive multiple sclerosis:

increasing clarity but many unanswered questions. J Neurol Sci 2002;

199:1–15.

[35] Hafler DA. The distinction blurs between an autoimmune versus

microbial hypothesis in multiple sclerosis. J Clin Invest 1999;104:

527–9.

[36] Luxton RW, Thompson EJ. Affinity distribution of antigen-specific

IgG in patients with multiple sclerosis and in patients with viral

encephalitis. J Immunol Methods 1990;131:277–82.