Journal of the Neurological Sciences, 122 (1994) 97-101
97
© 1994 Elsevier Science B.V. All rights reserved 0022-510X/94/$07.00 JNS 4228
Cerebrovascular and neurological disorders associated with antiphospolipid antibodies in CSF and serum Paolo Gallo *, Susanna Sivieri, Anna Maria Ferrarini, Bruno Giometto, Amelia Ruffatti a, Erika Ritter b Carlo Chizzolini b and Bruno Tavolato Institute of Neurology, Second Neurologic Clinic, and a Institute of Rheumatology, University of Padua, School of Medicine, via E. Vendramini, 7, 35137 Padua, Italy, and b Department of Immunology, FIDIA Research Laboratories, Abano Terme, Italy
(Received 25 June, 1993) (Revised, received 20 September, 1993) (Accepted 27 September, 1993) Key words: Antiphospholipid antibodies; Neurological diseases; Cerebrospinal fluid
Summary Paired serum and cerebrospinal fluid (CSF) samples from 70 patients with inflammatory and non-inflammatory neurological diseases, as well as 10 sera from patients with primary antiphospholipid syndrome (PAS), six of which presented with cerebrovascular ischemic syndromes, were studied for the presence of anticardiolipin antibodies (ACA) of the G and M classes. PAS sera and some selected paired CSF and serum specimens, were also analyzed for the presence of anti-phosphatidylserine (PS) and anti-phosphatidylethanolamine (PE) antibodies. High levels of IgG and IgM ACA were synthesized intrathecally only in patients with neurosyphilis. Patients with other infectious or inflammatory neurological diseases very rarely showed detectable levels of ACA in serum a n d / o r CSF. ACA were found not only in patients with untreated PAS but also in the serum of 3/7 patients with migraine, thus confirming a relationship between ACA and vascular disorders. The search for PS and PE antibodies disclosed that in PAS patients the serum titers of these antibodies mirrored ACA IgG and IgM titers, while they were never found in the CSF.
Introduction Antiphospholipid antibodies (aPLA), circulating polyclonal immunoglobulins of the three major classes (IgG, IgM and IgA), can be detected as anticardiolipin antibodies (ACA) in the serum of patients with systemic lupus erythematosus (SLE) and other autoimmune diseases (Harris et al. 1985; McNeil et al. 1991), and are clinically linked with the "primary antiphospholipid syndrome" (PAS), which features arterial and venous thrombosis, thrombocytopenia, and recurrentspontaneous abortion (Asherson et al. 1989). A wide spectrum of neurological disorders is thought to be associated with aPLA (Levine and Welch 1987; Briley et al. 1989): juvenile cerebrovascular ischemic syndromes (Brey et al. 1990; Levine et al. 1990; Hess et al. 1991), Sneddon's syndrome (Levine et al. 1988), Guillain-Barrd syndrome (GBS) (Harris et al. 1983), retroviral infections of the central nervous system (CNS), such as HTLV-I-associated myelopathy (Ges-
* Corresponding author. Tel.: 049.8216346; Fax: 049.8216358 SSDI 0022-510X(93)E0239- 6
sain et al. 1985) and AIDS (Canoso et al. 1987), multiple sclerosis (MS) (Lolli et al. 1991), and amyotrophic lateral sclerosis (ALS) (Meininger et al. 1992). A few studies have addressed the presence of aPLA in the cerebrospinal fluid (CSF) (Lolli et al. 1991; Marchiori et al. 1989) and it was speculated that these antibodies might play a role in the immunopathogenesis of CNS diseases. The aim of the present study was to extend these observations by investigating the presence of aPLA in paired CSF and serum samples from patients with various neurological diseases.
Materials and methods Patients
Paired serum and CSF specimens were obtained from the following groups of patients who were consecutively admitted to our clinic for diagnostic reasons: 30 with clinically definite MS (Poser et al. 1983) - all patients had intrathecal IgG synthesis as evaluated by the detection of IgG oligoclonal bands; eight suffered from neurosyphilis; seven f r o m e n c e p h a l i t i s / meningoencephalitis (two viral, three bacterial, one
98 tuberculous meningitis and one encephalo-meningeal sarcoidosis); nine from peripheral neuropathies (five axonal, two chronic inflammatory demyelinating polyneuropathy, two associated with monoclonal gammopathies); seven from non-inflammatory neurological diseases (three ALS, two Alzheimer, two ischemic ictus); two from stiff-man syndrome, and one from myasthenia gravis. CSF and serum were also obtained from four HIV-l-seropositive, neurologically asymptomatic children, as well as from a patient with transverse myelitis of unknown origin. The search for lupus anticoagulant (LAC) activity was not done in these patients. Also included in this study were serum samples from l0 patients with PAS (Hughes et al. 1986) who were also studied for the presence of LAC, two patients with SLE and aPLA, and seven patients with migraine. Among the 10 PAS patients, four had peripheral thrombosis or recurrent fetal loss, while six were referred for cerebrovascular ischemic syndromes. CSF and serum samples from 10 individuals with mild psychoneurotio disorders were used as controls (seven females, three males, mean age 38 years, range 18-45).
ELISA for PS and PE antibodies was basically performed as described (Umeda et al. 1989). Briefly, microplate wells were coated with 501) pmol PS or PE (FIDIA, Abano Terme), blocked with 3% bovine serum albumin in PBS for 2 h at room temperature, and then incubated overnight at 4°C with patient sera diluted 1:100. Bound antibody was detected with mouse antihuman IgG or lgM alkaline-conjugated antibodies. OD was determined at 405 nm using an ELISA reader, after addition of Sigma 104 phosphatase substrate; the cut-off value was 0.1.
Sample heating The presence of heat-sensitive masking cofactor(s) of aPLA has been described in normal human serum (Cheng and Wong 1990): IgG and IgA but not IgM reactivity to ACA increases when serum is heated to 56°C for 30 min (Cheng and Sam 1990; Cheng and Wong 1990). Since it is not known whether these inhibitory factors are also active at the CSF level, we studied paired CSF and serum specimens, before and after heating, from 10 patients with different ACA OD values, some of whom had intrathecat lgG synthesis (i.e., MS and neurosyphilis). Sera from four patients with PAS were also examined.
Routine CSF analysis Routine CSF analysis was performed as detailed previously (Gallo et al. 1988), and consisted in cell counting, calculation of IgG index and albumin C S F / serum ratio, detection of IgG oligoclonal bands by isoelectric focusing and specific immunofixation.
Statistics Student's t test for paired samples was applied; a P value lower than 0.05 was considered significant.
Results
Anti-cardiolipin, -phosphatidylserine (PS), -phosphatidylethanolamine (PE) antibody enzyme immunoassay (ELISA) The commercially available Elisa kit "Cardiolipin" (Kallestad, Chaska, MN, USA) was used according to the manufacturer's instructions. Serum was diluted 1:51 and CSF 1:2 in sample buffer. Preliminary experiments showed that the background was not influenced by the IgG content in either serum or CSF. Serum ACA were expressed in GPL and MPL units (corresponding to the cardiolipin binding activity of 1 /xg/ml of affinity purified IgG and IgM, respectively); conversion of OD values in GPL and MPL was also done according to the manufacturer's instructions; normal values are < 23 and < 11 for GPL and MPL, respectively. However, in order to compare our findings to previously published data, internal cut-off values were defined as mean + 3 × standard deviation (SD) of the OD of our normal control sera. CSF ACA were expressed as OD values; CSF OD higher than those of the corresponding serum were considered an expression of intrathecal ACA synthesis; the presence of a blood-brain barrier damage (i.e., increased albumin ratio) was always accurately considered.
The results of CSF and serum analysis for IgG and IgM ACA are summarized in Figs. 1 and 2. Four patients had more than 23 GPL in serum: one with neurosyphilis, the patient with a transverse myelitis of unknown etiology (enclosed in the encephalitis group, E, of Fig. 1), and two with migraine. Three patients,
~ MPL
GPI-
o
i
.
NC MS
NS
E
.
.
AD
.
"7"
P
. . . .
Nil)
i
M
NC MS ~lS
E
AD P
NI0 M
Fig. 1. G P L a n d M P L A C A levels in s e r u m . N C = n o r m a l c o n t r o l s ; M S = m u l t i p l e sclerosis; N S = n e u r o s y p h i l i s ; E = e n c e p h a l i t i s / m e n i n g o e n c e p h a l i t i s : • = b a c t e r i a l ; © = viral; • = s a r c o i d o s i s ; [] = myelitis o f u n k n o w n etiology; A D = a u t o i m m u n e d i s e a s e s : o = m y a s t e n i a gravis; • = s t i f f - m a n s y n d r o m e ; P = p o l y n e u r o p a t h i e s ; NID = non-inflammatory diseases; M = migraine; ....... cut-off suggested by Kallestad; ...... mean + 3 × SD of normal controls.
99 2-IOD
~" OD
if_ . . . . ~ ,
o-J
.-
.
AD
E
P
Nil)
+
:
NC MS NS AD
~
P
NID
Fig. 2. IgG and IgM A C A O D in the CSF. Abbreviations as in Fig. 1. ...... m e a n + 3 × SD of normal control OD.
two with neurosyphilis and one with migraine, had more than 11 MPL. Taking the mean + 3 × SD of the OD of the NC group as the cut-off value, five additional sera showed increased GPL values (> 16); two MS, one neurosyphilis, two polyneuropathies. No serum showed an MPL value between 11 and 5.6 (mean + 3 × SD MPL of NC group). CSF IgG ACA levels, expressed in OD, were significantly increased in four patients with neurosyphilis, in the patient with encephalomeningeal sarcoidosis (in two serial CSF) and, to a lesser degree, in only one MS patient. IgM OD was increased in four patients with neurosyphilis (two of whom also had high IgG OD) and, to a lesser degree, in one patient with MS (the same who showed an increased IgG OD). Three out of four sera, and all CSF from the HIV1-infected children, were positive for IgG ACA and negative for IgM ACA. However, these samples had been heated to 56°C for 30 min, in order to inactivate the virus. Before heating, only three sera had a GPL value > 16. After heating, a GPL value > 16 was found in 9/10 sera; this increase in GPL reactivity was statistically significant (P < 0.0001). Some CSF also showed increased IgG OD values after heating, in particular the samples from two patients with neurosyphilis and one with bacterial meningitis. Taken all together, the increase in GPL reactivity after heating was statistically significant also for the CSF (P = 0.041). As mentioned above, three patients with migraine showed increased serum ACA levels; two had > 23 GPL and one had > 11 MPL. All were females in the age range from 21 to 32 years. The first patient had classic migraine; headache in the second was accompanied by a transient neurologic deficit, while the third had a common migraine. All hematologic and immunologic tests, as well as all paraclinical examinations (evoked potentials, CT scan and MRI included), were normal in these subjects. After 3 months, two patients showed an increase in the ACA titer, while one became negative. None had a history of recurrent/ spontaneous abortion or thrombosis, and the platelet count was normal.
As expected, patients with PAS had the highest ACA values. After heating, the OD value for ACA IgG was remarkably increased, while ACA IgM were not influenced by heating (data not shown). To test whether cardiolipin was selectively recognized or whether other anionic phospholipids were also recognized in ACA-positive samples, reactivity to PS and PE was investigated. The detection of PS and PE antibodies in general reflected the presence of ACA. In particular, the highest GPL and MPL titers were found in PAS patients without neurological complications who also had the highest anti-PS and anti-PE IgG and IgM titers. Differences mostly concerned IgM in PAS patients with neurological complications where no MPL were detected, but PS and PE IgM titers were positive, although at a low level, in most samples. PS and PE antibodies were never detected in the CSF. Sera from three neurosyphilis and two MS patients showed low anti-PE IgM titers.
Discussion
The role of aPLA in the immunopathogenesis of autoimmune diseases, as well as the mechanism(s) by which they are involved in the pathogenesis of thrombosis, have not yet been clarified (Macworth-Young 1990). Also, no definitive conclusion on their role in inducing neurological disturbances by binding brain structures (e.g., neurons and myelin) can be drawn from the available literature (Harris and Hughes 1987). Our study shows that CSF IgG and IgM ACA levels are normal in most neurological patients, with the exception of individuals with neurosyphilis and encephalo-meningeal sarcoidosis who had intrathecally synthesized ACA. Lolli et al. (1991) found ACA in the CSF of 12/55 (22%) MS patients; however, 10 had a diagnosis of probable or possible MS (10/26, 38%), and only two were clinically definite MS (2/29, 7%). In our series of 30 clinically definite MS patients only one (3%) had ACA IgG in the CSF. The inclusion of patients with vascular (i.e., vasculitis) and other inflammatory/ infectious disorders (i.e., neuro-borreliosis) in the possible/probable MS group might possibly explain this discrepancy, as these conditions at onset may mimic MS both clinically and neuroradiologically. Some studies have described the detection of aPLA in the serum a n d / o r CSF of HIV-l-infected patients (Canoso et al. 1987; Cohen et al. 1986). It seems likely that sample heating may, at least in part, account for increased ACA levels in both the serum and CSF of these patients, as shown by our findings. The occurrence of aPLA in patients with cerebrovascular thrombosis has been widely investigated (Yagnik et al. 1984; Fisher et al. 1986; Briley et al.
100 1989; Chancellor et al. 1989; Brey et al. 1990; Levine et al. 1990; Hess et al. 1991). Our data further stress the relationship between increased ACA levels and the appearance of "vascular disorders". Interestingly, high cross-reactivity between cardiolipin and other anionic phospholipids, i.e. PS and PE, was observed in the PAS sera tested. This finding is in line with previous observations showing that thrombosis preferentially develops in patients where cross-reactivity can be demonstrated, while it is absent when such cross-reactivity is not observed (Harris et al. 1988). Briley et al. (1989) described the occurrence of headache as a prominent sign in six patients with aPLA, who also experienced important neurological complications; two had ischemic encephalopathy, two cerebral infarction, and two complicated migraine. Interestingly, only one of our three migraine patients with aPLA had a transient neurological deficit; the other two had uncomplicated migraine. At follow up, two had increased ACA, thus posing the question whether they had PAS (both were unmarried young women with no history of pregnancy). Migraine is usually considered an accessory symptom of PAS, but whether it constitutes an early manifestation of this syndrome deserves further investigation. A longitudinal study on the serum GPL and MPL levels in migraine patients is in progress. Although many case series suggested an association between aPLA and cerebral ischemia, no agreement exists regarding the presence of aPLA in unselected young populations with TIA/stroke, and published findings vary from 4% (Hart et al. 1984) to 45% (Brey et al. 1990). Our experience indicates that 10% of unselected (consecutive) TIA/ischemic stroke patients have increased ACA levels (data not shown). These patients, however, fell into a specific group; i.e., young adults, aged under 45, with no other risk factors for cerebrovascular diseases, some of whom will develop a PAS. Our findings indicate that individuals with elevated ACA fall into two neurological groups: (1) patients with infectious diseases, such as neurosyphitis and, in our study, encephalo-meningeal sarcoidosis, and (2) young patients with "vascular disorders", such as TIA/ischemic stroke and migraine. A search for ACA in other pathological conditions seems to lack clinical utility. Acknowledgements Supported by grants from Istituto Superiore di Sanit~ Progetto AIDS 1993 No. 8206-20.
References Asherson, R.A., Khamashta, M.A., Ordi-Ros, J., et al. (1989) The "primary" antiphospholipids syndrome: major clinical and serological features. Medicine, 68: 366-374.
Brey, R.L., Hart, R.G., Sherman, D.G. and Telegcr, C.[t. (1990) Anti-phospholipid antibodies and cerebral ischemia in young people. Neurology, 40:1190-1196. Briley, D.P., Coull, B.M. and Goodnight, S.H. (1989) Neurological disease associated with anti-phospholipid antibodies. Ann. New rol., 25: 221-227. Canoso, R.T., Zon, L.I. and Groopman, J.E. (1987)Anticardiolipin antibodies associated with HTLV-III infection. Br. J, Haematol., 65: 495-498. Chancellor, A.M., Glasgow, G.L., Ockelford, P.A., et al. (1989) Etiology, prognosis, and hemostatic function after cerebral infarction in young adults. Stroke, 20: 477-482. Cheng, H.-M. and Sam, C.-K. (1990) Bacterial immunity and immunogenesis of normal human salivary lgA and serum lgG anti-phospholipid antibody: a link? Immunol. Lett., 26: 7-10. Cheng, H.-M. and Wong, K.-K. (1990) Anti-phospho!ipid antibody isotypes in normal human sera. Immunol. Lett., 23: 183-186. Cohen, A.J., Phillips, T.M. and Kessler, C.M. (1986) Circulating coagulation inhibitors in the acquired immunodeficiency syndrome. Ann. Intern. Med., 104: 175-180. Del Papa, N., Meroni, P.L., Tincani, A., et al. (1992) Relationship between anti-phospholipid and anti-endothelial cell antibodies: further characterization of the reactivity on resting and cytokineactivated endothelial cells. Clin. Exp. Rheumatol., 10: 37-42. Fisher, M. and McGeehee, W. (1986) Cerebral infarct, TIA and lupus inhibitor. Neurology, 36: 1234-1237. Gallo, P., Bracco, F. and Tavolato, B. (1988) Blood-brain barrier damage restricts the reliability of quantitative formulae and isoelectric focusing in detecting intrathecally synthesized lgG.J. Neurol. Sci., 84: 87-93. Gessain, A., Barin, F., Gout, O., et al. (1985) Antibodies to human T-lymphotropic virus type-I in patients with tropical spastic paraparesis. Lancet, ii: 1247-1248. Harris, E.N. and Hughes, G.R.V. (1987) New reflections on cerebral lupus. In: J.A. Aarli, W.M.H. Behan and P.O. Behan (Eds.), Clinical Neuroimmunology, Blackwell Scientific Publ., London, pp. 340-353. Harris, E.N., Englert, H., Derue, G., et al. (1983) Anti-phospholipid antibodies in acute Guillain-Barr~ syndrome. Lancet, ii: 13611362. Harris, E.N., Gharavi, A.E. and Hughes, G.R.V. (1985) Anti-phospholipid antibodies. Clin. Rheum. Dis., 11: 591-610. Harris, E.N., Gharavy, A.E., Wasley, G.D. and Hughes, R.V. (1988) Use of an enzyme-linked immunosorbent assay and of inhibition studies to distinguish between antibodies to cardiolipin from patients with syphilis or autoimmune diseases. J. Infect, Dis., 157: 27-31. Hart, R.G., Miller, V.T., Coull, B,M. and Bril, V. (1984) Cerebral infarction associated with lupus anticoagulants: preliminary report. Stroke, 15: 114-118. Hess, D.C., Krauss, J., Adams, R.J., et al. (1991) Anticardiolipin antibodies: a study of frequency in TIA and stroke. Neurology, 41: 525-528. Hughes, G.R.V., Harris, E.N. and Gharavi, A.E. (1986) The anticardiolipin syndrome. J. Rheumatol., 13: 486-489. Levine, S.R. and Welch, K.M.A. (1987) The spectrum of neurologic disease associated with anti-phospholipid antibodies. Lupus anticoagulant and anticardiolipin antibodies. Arch. Neurol., 44: 876883. Levine, S.R., Langer, S.L., Albers, J.W. and Welch, K.M.A, (1988) Sneddon's syndromes: an antiphospholipid antibody syndrome? Neurology, 38: 798-800. Levine, S.R., Deegan, M.J., Futrell, N. and Welch, K.M.A. (1990) Cerebrovascular and neurologie disease associated with antiphospholipid antibodies: 48 cases. Neurology, 40: 1181-1189. Lolli, F., Matfi, S., Baruffi, M.C. and Amaducci, L. (1991) Cerebrospinal fluid anticardiolipin antibodies in neurological diseases. Clin. Immunol. Immunopathol., 59: 314-321.
101 Macworth-Young, C. (1990) Anticardiolipin antibodies: more than just a disease marker. Immunol. Today, 11: 60-65. Marchiori P.E, Reis, M., Quevedo, M.E.Z., et al. (1989) Anti-phospholipid antibodies in some neuroimmunological diseases. In: 7th International Congress of Immunology, Berlin, July 30-August 5. Abstract Book (Abstract 87-18). McNeil, H.P., Chesterman, C.N. and Krilis, S.A. (1991) Immunology and clinical importance of antiphospholipid antibodies. Adv. Immunol., 49: 193-280. Meininger, V., Rozier, A., Dib, M., et al. (1992) Anti-phospholipid antibodies in amyotrophic lateral sclerosis: an analysis in 409 patients. Neurology, 42 (Suppl. 3): 202.
Poser, C.M., Paty, D.W., Scheinberg, L., et al. (1983) New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann. Neurol., 13: 227-231. Umeda, M., Igarashi, K., Nam, K.S. and Inoue, K. (1989) Effective production of monoclonal antibodies against phosphatidylse$ine: Stereo-specific recognition of phosphatidylserine by monoclonal antibody. J. Immunol., 142: 2273-2279. Yagnik, P. and Dhophes, V. (1984) Stroke and circulating lupus. Stroke, 15: 388-389.