Cardiolipin and β₂-Glycoprotein I antibodies associate with cognitive impairment and seizure frequency in developmental disorders.
ABSTRACT Cardiolipin (CL) and β(2)-Glycoprotein I (β(2)-GpI) antibodies have been shown to associate with various neurological symptoms including seizures and cognitive dysfunction. Here we studied the prevalence of CL, β(2)-GpI and antinuclear (ANA) antibodies in 74 patients with various developmental disorders with epilepsy and 70 healthy controls. Developmental disorders were classified into genetic syndromes and diseases, genetic and/or acquired conditions, cortical dysgenesias and acquired encephalopathias. IgM-CL and β(2)-GpI antibodies were significantly more common in patients (46% vs. 20%, p<0.001 and 10% vs. 0%, p<0.05). Patients with most frequent seizures were more likely to have IgM-CL antibodies. The risk for positive IgM-CL, IgG-CL and β(2)-GpI antibodies increased concomitantly with increasing intellectual disability. Present data demonstrates that epilepsy with frequently recurring seizures may be associated with secondary immune system activation.
- [Show abstract] [Hide abstract]
ABSTRACT: Over the last few years autoantibodies against neuronal proteins have been identified in several forms of autoimmune encephalitis and epilepsy. NMDA receptor (NMDAR) and voltage gated potassium channel (VGKC) complex antibodies are mainly associated with limbic encephalitis (LE) whereas glutamic acid decarboxylase antibodies (GADA) and anticardiolipin (ACL) antibodies are more commonly detected in patients with chronic epilepsy. Clinical features vary between these antibodies suggesting the specificity of different neuronal antibodies in seizures. Serum samples of 14 GADA positive and 24 ACL positive patients with refractory epilepsy were analyzed for the presence of VGKC or NMDAR antibodies. No positive VGKC or NMDAR antibodies were found in these patients. The results confirm the different significance of these neuronal antibodies in seizure disorders. Different autoantibodies have different significance in seizures and probably have different pathophysiological mechanisms of actions.Epilepsy research 01/2014; · 2.48 Impact Factor
Cardiolipin and b2-Glycoprotein I antibodies associate with cognitive
impairment and seizure frequency in developmental disorders
K.A. Lehtima ¨kia,*, J. Peltolab, S. Liimatainenb, A.-M. Haapalac, M. Arviod
aDepartment of Neurosurgery, Tampere University Hospital, Teiskontie 35, 33521 Tampere, Finland
bDepartment of Neurology and Rehabilitation, Tampere University Hospital, Finland
cCentre of Laboratory Medicine, Clinical Microbiology, Tampere University Hospital, Finland
dDepartment of Child Neurology, Pa ¨ija ¨t-Ha ¨me Central Hospital, Lahti, Finland
Phospholipid (PL) antibodies are a heterogeneous group of
autoantibodies binding to phospholipid-like antigens like cardi-
olipin (CL)and phosphatidylserine(PS).b2-GlycoproteinI(b2-GpI)
is a plasma protein binding to negatively charged macromolecules
and structures, such as phospholipids, DNA, platelets and
mitochodria.1b2-GpI mRNA is also expressed in neurons,
astrocytes and in brain endothelial cells.2Phospholipid and b2-
GpI antibodies have reported to bind to neuronal tissue.3–6
High systemic titres of these antibodies are characteristic to
primary antiphospholipid syndrome (APS), with systemic mani-
festations of recurrent spontaneous pregnancy loss, thrombosis,
symptoms.7These neurological symptoms include chorea, psy-
are relatively well characterized neurological symptoms associat-
ed with PL antibodies, especially in patients with systemic lupus
erythematosus (SLE).9,10However, PL antibodies have been found
in patients with epilepsy without SLE.11–13
In the present study we aimed to study the role of CL, b2-GpI
and antinuclear (ANA) autoantibodies in different types of
developmental disorders. We have especially focused on whether
these antibodies are associated with the level of cognitive
dysfunction, etiological factors underlying intellectual disability
and epilepsy, specific epilepsy syndromes or frequency of seizures.
2. Materials and methods
The inclusion criteria for the study were a developmental
disorder due to a genetic or structural cause with at least five-year
history of refractory seizures despite of adequate treatment. Of 77
consecutively recruited patients, written informed consent was
obtained from guardians of 74 patients. A total of 70 healthy blood
donors were used as control subjects.
laboratory tests. We used commercially available assays to
determine CL antibody titres (Quanta Lite ACA IgG and IgM; B2
GPI, IgG; INOVA Diagnostics, San Diego, CA, USA). Patients and
controls with positive results in IgG-CL assay were also tested for
distinct IgG class b2-GpI antibodies. Antinuclear (ANA) antibodies
were measured indirectly using Hep-2 cells as antigens (INOVA
Diagnostics, San Diego, CA, USA), and presence of antibodies were
determined using fluoromicroscopy. The limit values between
positive and negative results were 12.5 international MPL units for
IgM-CL antibodies; 15 international GPL units for IgG-CL anti-
bodies, and 20 international SGU units for b2-GpI antibodies.
Seizure 20 (2011) 438–441
A R T I C L EI N F O
Received 29 November 2010
Received in revised form 22 January 2011
Accepted 31 January 2011
A B S T R A C T
Cardiolipin (CL) and b2-Glycoprotein I (b2-GpI) antibodies have been shown to associate with various
neurological symptoms including seizures and cognitive dysfunction. Here we studied the prevalence of
CL, b2-GpI and antinuclear (ANA) antibodies in 74 patients with various developmental disorders with
epilepsy and 70 healthy controls. Developmental disorders were classified into genetic syndromes and
diseases, genetic and/or acquired conditions, cortical dysgenesias and acquired encephalopathias. IgM-
CLandb2-GpIantibodiesweresignificantlymorecommoninpatients(46%vs.20%,p < 0.001and10%vs.
0%, p < 0.05). Patients with most frequent seizures were more likely to have IgM-CL antibodies. The risk
for positive IgM-CL, IgG-CL and b2-GpI antibodies increased concomitantly with increasing intellectual
disability. Present data demonstrates that epilepsy with frequently recurring seizures may be associated
with secondary immune system activation.
? 2011 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: +358 331169412; fax: +358 331164373.
E-mail addresses: email@example.com (K.A. Lehtima ¨ki), firstname.lastname@example.org
(J. Peltola), email@example.com (S. Liimatainen),
firstname.lastname@example.org (A.-M. Haapala), email@example.com (M. Arvio).
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/yseiz
1059-1311/$ – see front matter ? 2011 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved.
Pearson Chi-square test, Fisher’s exact test and logistic
regression analysis served in statistical analysis. A p value of
0.05 or less was considered statistically significant. The ethical
committee of Tampere University Hospital approved the study
3.1. Clinical data
The study group comprised of 74 patients (31 females and 43
males, age range 15–61, mean age 35 years) and 70 controls (age
range 16–65 years, mean age 36 years). All patients had undergone
a detailed analysis of their condition including video-EEG, brain
The patients’ etiological diagnoses in detail (classified into genetic
syndromes, genetic diseases, genetic and/or acquired conditions,
cortical dysgenesias and acquired disorders)are presentedin Table
1 and epilepsy diagnoses in Table 2. Sixty-four patients (86%)
experienced first seizuresduring childhoodor adolescence.Twelve
patients showed hippocampal sclerosis (eight had an acquired
encephalopathia and four a genetic disorder) in brain MRI. Two
patients acted at subnormal intelligence, 7 at the level of mild (IQ
69–50), 4 at moderate (IQ 49–35), 30 at severe (IQ 34–20), and 31
at the level of profound (IQ < 20) intellectual disability (ID). Nine
patients suffered from 1 to 11 seizures yearly, 27 from 1 to 3
seizures monthly, 21 from 1 to 6 seizures weekly, and 17 patients
from daily seizures.
All patients except one were on polytherapy. Fifty-five patients
received valproate, 28 topiramate, 25 lamotrigine, 16 oxcarbatse-
pine, 10 levetirasetame, 9 carbamazepine, 9 clobatsame, 8
clonazepam, 7 vigabatrine, 6 gabapentine, 5 phenytoin, 2
ethosuximide, 2 azetatsolamide, and 2 phenobarbitale.
3.2. Antibody tests
3.2.1. Patients and controls
Positive titres of IgM-CL antibodies were more frequent in
patients (46%) than in controls (20%) (p < 0.001; Fisher’s exact test,
Table 2). The prevalence of positiveIgG-CL antibodies did not differ
between patients (16%) and controls (13%), however, 10% of
patients had b2-GpI antibodies while no b2-GpI antibodies were
detected in controls (p < 0.05; Fisher’s exact test, Table 2). The
prevalence of ANA antibodies was the same among patients and
controls. The patient’s age, gender, onset age of epilepsy or
medication did not relate to the prevalence of PL or ANA
3.2.2. Etiology of developmental disorder
The prevalence of positive antibody titres in different etiologi-
cal groups is presentedin Table 2. Autoantibodies were not specific
to any of the etiological groups, but were distributed within all
groups studied. The prevalence of positive antibodies was highly
dependent of the number of patients in each group (Table 2).
3.2.3. Epilepsy syndrome and seizure frequency
Autoantibodies studied were not specific to any of the epileptic
syndromes. There were no statistically significant differences
between epileptic syndromes, although IgM-CL antibodies were
most frequent in Lennox-Gastaut syndrome (Table 2). Prevalence
of antibodies was highly dependent of the group size. We found no
association between positive antibodies and presence of hippo-
Etiological groups and specific diagnoses of the 74 study patients.
Etiological groups Diagnosis
Genetic syndromes (15) Down syndrome (4), X-linked mental retardation (3), trisomy 15q (2), Angelman (1), deletion 6q (1), autosomal dominant
familial disorder (1), Kabuki (1), unidentified dysmorphic syndrome (2)
Aspartylglucosaminuria (2), spinocerebellar ataxia (1), neurofibromatosis type 2 (1), MELAS (1), x-linked familial (1),
Infantile autism (4), unspecific intellectual disability with generalized epilepsy (2)
Polymicrogyria (6), temporal dysmyelination (4), microcephalia (2), schizenkefalia (1), pachygyria (1), lissenencephalia (1)
post-asphyxial (11), post-infectious (8), mesial temporal sclerosis (5), early severe epilepsy (4) and post-traumatic (3)
Genetic diseases (7)
Genetic/acquired condition (6)
Cortical dysgenesias (15)
Acquired encephalopathias (31)
Prevalence of positive antibodies in controls and patients according to etiology and epilepsy syndrome.
IgM-CL (%)IgG-CL (%)
b2-GpI (%)ANA (%)
All patients (n=74)
Genetic syndrome (n=15)
Genetic disease (n=7)
Genetic/acquired condition (n=6)
Cortical dysgenesias (n=15)
Acquired encephalopathia (n=31)
Lennox-Gastaut syndrome (n=18)
Generalized epilepsy (degenerative) (n=6)
Generalized epilepsy (idiopathic) (n=5)
TLE with HS (n=12)
Extra-TLE focal/multifocal epilepsy (n=26)
Reflex epilepsy (n=1)
Infantile spasms (n=1)
Dravet syndrome (n=1)
Abbreviations: Ab, antibody; ANA, antinuclear antibody; CL, cardiolipin; CSWSS, continuous spike-waves during slow-wave sleep epilepsy syndrome; Gp, glycoprotein; HS,
hippocampal sclerosis; IgG, immunoglobulin class G; IgM, immunoglobulin class M; TLE, temporal lobe epilepsy.
Statistical difference between patients and controls: *** (p<0,001), and * (p<0,05; Fisher’s exact).
K.A. Lehtima ¨ki et al./Seizure 20 (2011) 438–441
campal sclerosis on MRI. Prevalence of IgM-CL antibodies
increased concomitantly with increasing seizure frequency com-
pared to controls (Table 3). b2-GpI and ANA antibodies were not
significantly associated to seizure frequency.
3.2.4. Intellectual disability
disability, subnormal, mild and moderate levels of disability were
pooled as a single group (IQ > 35; n = 13), severe as one group (IQ
20–34; n = 30) and profound disability as one group (IQ < 20;
n = 31).TheprevalenceofpositiveIgM-CLantibodiesincreasedwith
increasing intellectual disability (Table 4). IgG-CL antibodies were
also more frequent in profound intellectual disability compared
with controls (Table 4). The prevalence of b2-GpI antibodies
increased significantly with increasing intellectual disability (Table
cognitive impairment (data not shown).
3.2.5. Antibody titres
Among 34 IgM-CL positive patients 15 (44%) were regarded as
moderately positive (21–80 MPL units) and within 14 control
subjects 5 (36%) were moderately positive. Other positive patients
and controlsubjects had low positivetitres. Among IgG-CL positive
patients 7 out of 12 (58%) were regarded as moderately positive
(21–80 GPL units), however, in controls, IgG-CL was moderately
subjects had low positive titres. Highest detected CL titres
(moderate) were not found to be associated to any specific
diagnosis or etiology. Observed titres for b2-GpI antibodies were
between 39 and 82 international SGU units.
In the present study we found that in developmental disorders
manifesting with epilepsy the prevalence of IgM-CL and b2-GpI
antibodies is significantly higher compared to healthy blood
donors. Interestingly, increase in seizure frequency increased
concomitantly the risk for positive IgM-CL antibodies. Finally, we
found that increasing severity of intellectual disability increased
significantly the risk for positive IgM-CL, IgG-CL and b2-GpI
Previously, IgG-CL antibodies were found in patients with
recent seizures, but not in patients who were seizure free for a
month prior to sampling.13In a large population based study by
Ranua et al. authors described that patients with partial epilepsy
with at least one seizure per month were more likely to have
positive IgG-CL antibodies than patients with lower seizure
frequency.14The authors of these studies suggested that produc-
tion of PL antibodies reflects secondary immune system activation
related to seizures. Our results further support this hypothesis.
Whether autoantibodies areassociated with the developmental
disorders or epilepsy as such cannot be fully evaluated in the
present study since patients with developmental disorder without
epilepsy were not included in the study. However, the prevalence
of CL and b2-GpI antibodies in our patients with developmental
disorder and refractory epilepsy is in accordance with a study in
adult patients with localization related epilepsy.11We therefore
conclude that epilepsy and seizures is more clearly associated with
PL antibodies than developmental disorders as per se.
The pathophysiological relevance of PL antibodies in develop-
mental disorders and refractory epilepsy is controversial. In SLE,
moderate or high titres of CL-IgG are associated with seizures.9SLE
patients with positive IgG-CL antibodies have 3.7 times higher risk
for seizures compared to CL antibody negative patients,10
suggesting that antibodies may indeed decrease seizure threshold
level. PL antibodies have been shown to depolarize neurons in
vitro.15Moreover, CL antibodies have been described to reduce
gamma-aminobutyric acid (GABA) mediated inhibition of neu-
rons.16PL antibodies have suggested to cause epilepsy in a few
Impairment of cognitive functions in SLE patients with PL-
antibodies is well documented.8We found that the prevalence of
The prevalence of CL and b2-GpI antibodies in patients and controls according to seizure frequency.
Pos (%)OR (95%CI)
Pos (%)OR (95%CI)
All patients (n=74)
3.4 (1.6–7.1) 0.00112 (16)
1.3 (0.5–3.3)0.577 (10)*
Statistical difference between particular patient group and control group was evaluated using logistic regression analysis (IgM-CL and IgG-CL). Abbreviations: b2-GpI, beta2
glycoprotein I; CI, confidential interval; CL, cardiolipin; IgG, immunoglobulin class G; IgM, immunoglobulin class M; NC, not calculable; OR, odds ratio.
Statistical difference between patients and controls: *, p<0,05; Fisher’s exact).
The prevalence of CL and b2-GpI antibodies in patient groups according to the severity of intellectual disability.
Pos (%)OR (95%CI)
Pos (%) OR (95%CI)
Subnormal–moderate (IQ>35, n=13)
Severe (IQ 20–34, n=30)
Profound (IQ<20, n=31)
All patients (n=74)
3.4 (1.6–7.1)0.00112 (16)
1.3 (0.5–3.3)0.577 (10)*
Statistical difference between particular patient group and control group was tested using logistic regression analysis (IgM-CL and IgG-CL). Abbreviations: b2-GpI, beta2
glycoprotein I; CI, confidential interval; CL, cardiolipin; IgG, immunoglobulin class G; IgM, immunoglobulin class M; OR, odds ratio.
Statistical difference between patients and controls: *, p<0,05 (Fisher’s exact). Statistical difference between patient groups:y, p<0,05 (qhi square test).
K.A. Lehtima ¨ki et al./Seizure 20 (2011) 438–441
IgG/M-CL andb2-GpI antibodies was considerably high in patients
with most severe forms of intellectual disability compared to
tissue injury manifesting with cognitive impairment. However,
further studies are needed to confirm our findings.
Several factors limit the interpretation of our data. Classifica-
tion of patients into smaller heterogeneously sized subgroups
limits the evaluation of antibody prevalence, especially if antibody
prevalence is low (Ig-CL, b2-GpI and ANA). IgM-CL had clearly
highest prevalence both in healthy controls and all patient groups,
enabling most reliable calculations. However, the pathophysiolog-
ical relevance of IgM-CL antibodies in epilepsy is entirely
unknown, and is most likely limited due to its relatively high
prevalence in healthy controls. However, we suggest that
increased IgM-CL prevalence in patients with frequent seizures
reflects host’s non-specific immunological activation as a response
to frequent seizures.
As a conclusion, our results as well as previous data indicate
that developmental disorders
associatedwithlow tomoderatetitresof CLandb2-GpI antibodies.
High seizure frequency and especially severe intellectual disability
were significantly associated to positive antibody titres. Observed
immunological activation is most likely secondary phenomenon
due to high seizure frequency and neuronal tissue damage.
with refractory epilepsyare
Conflict of interest
Authors did not have any conflicts of interests.
Study was financially supported by competitive EVO funding of
Pirkanmaa Hospital District (neurosurgery). We thank Heini
Huhtala (Department of Health Sciences, University of Tampere)
for the assistance in statistical analysis.
1. Kandiah DA, Krilis SA. Beta 2-glycoprotein I. Lupus 1994;3:207–12.
2. Caronti B, Calderaro C, Alessandri C, Conti F, Tinghino R, Palladini G, Valesini G.
Beta2-glycoprotein I (beta2-GPI) mRNA is expressed by several cell types
involved in anti-phospholipid syndrome-related tissue damage. Clinical and
Experimental Immunology 1999;115:214–9.
3. Sun KH, Liu WT, Tsai CY, Liao TS, Lin WM, Yu CL. Inhibition of astrocyte
proliferation and binding to brain tissue of anticardiolipin antibodies purified
from lupus serum. Annals of the Rheumatic Diseases 1992;51:707–12.
4. Kent M, Alvarez F, Vogt E, Fyffe R, Ng AK, Rote N. Monoclonal antiphosphati-
dylserine antibodies react directly with feline and murine central nervous
system. The Journal of Rheumatology 1997;24:1725–33.
5. Kent MN, Alvarez FJ, Ng AK, Rote NS. Ultrastructural localization of monoclonal
antiphospholipid antibody binding to rat brain. Experimental Neurology
bind tocentral nervous system.
7. Levine SR, Welch KM. The spectrum of neurologic disease associated with
antiphospholipid antibodies. Lupus anticoagulants and anticardiolipin antibo-
dies. Archives of Neurology 1987;44:876–83.
8. Sanna G, Bertolaccini ML, Cuadrado MJ, Laing H, Khamashta MA, Mathieu A,
Hughes GR. Neuropsychiatric manifestations in systemic lupus erythematosus:
prevalence and association with antiphospholipid antibodies. The Journal of
9. Herranz MT, Rivier G, Khamashta MA, Blaser KU, Hughes GR. Association
between antiphospholipid antibodies and epilepsy in patients with systemic
lupus erythematosus. Arthritis and Rheumatism 1994;37:568–71.
10. Liou HH, Wang CR, Chen CJ, Chen RC, Chuang CY, Chiang IP, Tsai MC. Elevated
levels of anticardiolipin antibodies and epilepsy in lupus patients. Lupus
11. Peltola JT, Haapala A, Isoja ¨rvi JI, Auvinen A, Palmio J, Latvala K, Kulmala P, Laine
S, Vaarala O, Kera ¨nen T. Antiphospholipid and antinuclear antibodies in
patients with epilepsy or new-onset seizure disorders. American Journal of
12. Eriksson K, Peltola J, Kera ¨nen T, Haapala AM, Koivikko M. High prevalence of
antiphospholipid antibodies in children with epilepsy: a controlled study of 50
cases. Epilepsy Research 2001;46:129–37.
13. Liimatainen S, Peltola M, Fallah M, Kharazmi E, Haapala AM, Peltola J. The high
prevalence of antiphospholipid antibodies in refractory focal epilepsy is related
to recurrent seizures. European Journal of Neurology 2009;16:134–41.
14. Ranua J, Luoma K, Peltola J, Haapala AM, Raitanen J, Auvinen A, Isoja ¨rvi J.
Anticardiolipin and antinuclear antibodies in epilepsy—a population-based
cross-sectional study. Epilepsy Research 2004;58:13–8.
15. Chapman J, Cohen-Armon M, Shoenfeld Y, Korczyn AD. Antiphospholipid
antibodies permeabilize and depolarize brain synaptoneurosomes. Lupus
16. Liou HH, Wang CR, Chou HC, Arvanov VL, Chen RC, Chang YC, Chuang CY, Chen
CY, Tsai MC. Anticardiolipin antisera from lupus patients with seizures reduce a
GABA receptor-mediated chloride current in snail neurons. Life Science
17. Angelini L, Granata T, Zibordi F, Binelli S, Zorzi G, Besana C. Partial seizures
associated with antiphospholipid antibodies in childhood. Neuropediatrics
K.A. Lehtima ¨ki et al./Seizure 20 (2011) 438–441