Clinical manifestations of chronicinflammatory demyelinating polyneuropathywith anti-cardiolipin antibodies

Chronic inflammatory demyelinating polyneurop-athy (CIDP) and Guillain–Barre syndrome (GBS)are autoimmune disorders of the peripheral ner-vous system characterized by elevated cerebrospi-nal fluid (CSF) protein concentrations. Althoughbiopsy specimens of CIDP exhibit demyelinationor endoneural edema similar to findings in GBS,the diseases are considered separate clinical entities(1–3) because of differences in clinical course andresponse to immunosuppressive treatment. Anti-ganglioside antibodies have been detected in serafrom patients with GBS, where several specificantibodies appear related to clinical variables (4–6). Humoral immunity therefore plays a role in thepathogenesis of GBS. In patients with CIDP,however, prevalence of antibodies against notonly glycolipids but also myelin antigens was notsignificantly elevated beyond those in a controlgroup (1). Clinically, CIDP has been accompaniedby several systemic diseases, such as systemic lupus

erythmatosus (SLE) (7), human immunodeficiencyvirus infection (8), diabetes mellitus (9), andmonoclonal gammopathy (10). These findingssuggest that CIDP has a heterogeneous pathogen-esis or is merely a clinical syndrome that can resultfrom several pathogenic mechanisms.Previous pathologic and immunologic studies

have demonstrated that both humoral and cell-mediated immunity are likely to contribute to thepathogenesis of CIDP (11, 12). As humoralimmune factors, a small group of patients withCIDP presenting with motor-predominant neur-opathy display anti-GM1 antibodies in sera (13),while IgM anti-GM1 antibodies are frequentlydetected in patients with multifocal motor neurop-athy. Moreover, patients with CIDP and serumIgM monoclonal gammopathy present with chro-nic sensory-predominant neuropathy and fre-quently demonstrate anti-MAG/SGPG antibodies(14). These findings suggest that humoral

Acta Neurol Scand 2005: 111: 258–263 DOI: 10.1111/j.1600-0404.2005.00387.x Copyright � Blackwell Munksgaard 2005

ACTA NEUROLOGICASCANDINAVICA

Nakajima H, Shinoda K, Doi Y, Tagami M, Furutama D, Sugino M,Kimura F, Hanafusa T. Clinical manifestations of chronic inflammatorydemyelinating polyneuropathy with anti-cardiolipin antibodies.Acta Neurol Scand 2005: 111: 258–263. � Blackwell Munksgaard 2005.

Objective – Chronic inflammatory demyelinating polyneuropathy(CIDP) is an autoimmune syndrome where certain autoantibodiesdefine clinicopathologic subgroups. In the present study, serum anti-cardiolipin antibodies (aCL) were evaluated. Materials and methods –We investigated aCL in sera from 21 patients diagnosed with CIDP inour hospital between 1991 and 2001. The four CIDP patients with aCL(aCL+) were compared with 17 patients without aCL (aCL)).Results – All aCL+ patients displayed sensory-motor polyneuropathy,with severity and distribution of weakness resembling those in aCL)patients. Anti-nuclear antibody titer of aCL+ patients weresignificantly higher than those in aCL) patients. None of aCL+patients presented clinical manifestations of primary anti-phospholipidsyndrome (APS), such as thromboses or recurrent abortion. Althoughthe aCL+ patients were older and had more complications and moresevere pathologic features than aCL) patients, they responded well tosteroid pulse or intravenous immunoglobulin. Conclusion – The aCLin CIDP apparently differ from �autoimmune� aCL in APS, instead beingamong the autoantibodies pathologically involved in CIDP subgroups.

H. Nakajima1, K. Shinoda2, Y. Doi1,M. Tagami1, D. Furutama1,M. Sugino1, F. Kimura1,T. Hanafusa11Division of Neurology, First Department of InternalMedicine, Osaka Medical College, Osaka, Japan;2Department of Internal Medicine, Hanwa SumiyosiHospital, Hanwa, Japan

Key words: chronic inflammatory demyelinatingpolyneuropathy; anti-cardiolipin antibody; intravenousimmunoglobulin; steroid therapy

Hideto Nakajima, Division of Neurology,First Department of Internal Medicine, Osaka MedicalCollege, 2-7 Daigakumachi, Takatsuki City, Osaka 569-8686, JapanTel.: +81 72 683 1221Fax: +81 72 683 1801email: in1045@poh.osaka-med.ac.jp

Accepted for publication November 19, 2004

258

immunity plays an important role in some patientswith CIDP, and that some specific autoantibodiesmay contribute to subgroups of CIDP.Anti-cardiolipin antibodies (aCL) and lupus

anticoagulant are both anti-phospholipid antibod-ies. The presence of these antibodies is associatedwith arterial or venous thrombosis and recurrentabortion or fetal loss (15). These clinical manifes-tations accompanied by the presence of anti-phospholipid antibodies are recognized as theanti-phospholipid syndrome (APS). Serum frompatients with SLE or other autoimmune diseasesfrequently displays aCL, and these antibodies arealso reportedly associated with various neurologicdiseases including stroke, transverse myelopathy,migraine, epilepsy, chorea and GBS (15, 16).However, an association between aCL and CIDPhas yet to be elucidated.The present study investigated aCL in sera from

21 patients diagnosed with CIDP. The four CIDPpatients with aCL were compared with patientswithout aCL.

Materials and methods

Patients

Twenty-one patients diagnosed with CIDP whowere examined at our institution between 1991 and2001 were screened for aCL. All patients under-went clinical, laboratory, electrophysiologic andpathologic studies, and fulfilled the publishedcriteria for idiopathic CIDP (17). Of the21 patients, four had detectable IgG aCL elevation(aCL+). Features in these patients were reviewedto identify any distinct clinical, neurologic orlaboratory features. To evaluate the disability ofthe neuropathy on daily functional activities, themodified Rankin Scale Score was used. The gradesof this scale range from 0 to 5 as follows:0, asymptomatic; 1, non-disabling symptoms thatdo not interfere with daily activities; 2, slightdisability, unable to carry out all activities but stillable to look after oneself; 3, moderate disability,requiring assistance with some activities but able towalk without assistance; 4, moderately severedisability, unable to walk without assistance andunable to attend to one’s own bodily needs withoutassistance; and 5, severe disability, totally depend-ent, requiring constant nursing care and attention.

Laboratory studies

IgG aCL, IgM aCL and b2-glycoprotein I(b2-GPI) were measured using standardizedELISA (SRL, Inc., Tokyo, Japan). Anti-nuclear

antibodies (ANA), anti-double-stranded DNAantibodies (anti-dsDNA), anti-RNP antibodies(anti-RNP), anti-Sm antibodies (anti-Sm), andanti-SSA/B antibodies (anti-SSA/SSB) were detec-ted by ELISA (Department of Central Laboratory,Osaka Medical College, Osaka, Japan). Anti-ganglioside antibodies (anti-GM1, anti-GD1a,anti-GD1b and anti-GQ1b) were measured by amodified ELISA (18).

Electrophysiologic investigation

All patients were investigated using motor nerveconduction and F-wave studies of the median,ulnar, deep peroneal, and tibial nerves. Antidromicsensory conduction studies were performed usingdistal stimulation of the median, ulnar, and suralnerves. Concentric-needle electromyography wasperformed in the first dorsal interosseous andtibialis anterior muscles. Amplitude, area, andduration of the negative peak of the compoundmuscle action potential (CMAP) and amplitude ofthe sensory nerve action potential (SNAP) weremeasured. Conduction block was defined as areduction of proximal CMAP compared with distal(P/D) exceeding 50%. Increased temporal disper-sion was defined as greater than 30% prolongationof CMAP duration (P/D). These findings weretaken as evidence of demyelination. Prior to theinvestigation, tested limbs were warmed in water at37�C.

Histopathologic examination of nerve

Sural nerves specimens were obtained under localanesthesia by biopsy at the ankle above the lateralmalleolus. A portion of the nerve tissue was fixedin 2% buffered glutaraldehyde, post-fixed in 1%buffered osmium tetroxide, and dehydrated inacetone. The material was impregnated with andembedded in Epon, which was allowed to poly-merize at 60�C for 24 h. Transverse sections of theentire sural nerve, 1 lm in thickness were stainedwith alkaline toluidine blue or 1% p-phenylene-diamine. For teased fiber preparations, nerve tissuewas fixed in 2% buffered glutaraldehyde, post-fixedin 1% buffered osmium tetroxide, dehydrated inacetone, and impregnated with Epon. Nerve fiberswere then teased in a film of Epon on a microscopeslide.

Statistical analysis

Differences of the median levels between aCL+and aCL) groups were tested using the Mann–Whitney U-test. Within-patient levels of aCL

CIDP with anti-cardiolipin antibodies

259

before and after treatment were compared usingthe Wilcoxon signed-rank test for matched pairsdata. P-values <0.05 were considered to bestatistically significant.

Results

Clinical features in aCL-positive patients

IgG aCL were found in four patients. The aCL+patients tended to be somewhat older than aCL)patients but this difference did not reach a level ofstatistical significance (54 vs 46). Severity anddistribution of weakness, and gait disturbance weresimilar between aCL+ and aCL) patients.Although all aCL+ patients showed sensorysymptoms and findings, four of the 17 aCL)patients displayed pure motor neuropathy, threeaCL) patients showed motor-dominant neuro-pathy. Most patients of both groups experienceda monophasic course or relapse. However, three ofthe aCL) patients experienced a chronic course(Table 1).Table 2 showed the clinical features of four

aCL+ patients, no patients had detectable IgMaCL or b2-GPI. One patient was treated for CIDPwith intravenous immunoglobulin (IVIG), whileothers were treated with steroid pulse therapy. AllaCL+ patients responded well to these treatments.

Two of these patients achieved remission.Although the remaining two patients demonstrateda relapsing–remitting course, significant improve-ment on functional scales was observed with eachcourse of steroid pulse therapy. As compared withaCL) patients, the degree of improvement ofRankin score of the aCL+ patients was moreremarkable (2.75 vs 2.00, P ¼ 0.03). All aCL+patients had either diabetes mellitus, lung cancer,myocardial infarction, or alcoholism as concurrentillness, but none of them showed manifestations ofprimary APS such as thrombotic events, recurrentabortion, pulmonary hypertension, or chorea.

Laboratory, electrophysiologic, and pathologic features inaCL-positive patients

ANA was detected in all four aCL+ patients, andthe mean ANA titer tended to be somewhat higherthan in aCL) patients but this difference did notreach a level of statistical significance (1:1520 vs1:36) (Table 3). On the contrary, antibodies relatedto specific collagen diseases such as anti-dsDNA,anti-RNP, anti-Sm, and anti-SSA/SSB were notdetected. Anti-ganglioside antibodies were alsomeasured using ELISA. IgG anti-GM1 antibodieswere not detected in patients with aCL. Of the17 aCL) patients, two had IgG anti-GM1, IgGanti-GD1a, or IgG anti-GD1b antibodies. Noassociations were identified between presence ofaCL and pattern of anti-ganglioside antibodies.In electrophysiologic studies, no inter-group

differences were noted in median and ulnar nerveconduction abnormalities such as conductionblock, delayed conduction velocities, or prolongeddistal latencies. In tibial nerve, however, aCL+patients showed significant prolonged distallatency and delayed conduction velocity as

Table 1 Clinical features of 21 patients with CIDP

Patients withaCL (n ¼ 4)

Patients withoutaCL (n ¼ 17)

Age (years), mean (range) 54 (30–62) 46 (23–83)M/F 2/2 9/8Duration (months), mean (SD) 59 (50) 72 (65)Neurological symptom, SM/M/MD 4/0/0 10/4/3Rankin score, 2/3/4/5 0/2/1/1 4/8/4/1Therapy, CS (SP)/PP/IVIG 3(3)/0/1 6(2)/7/4Course, monophasic/relapsing/chronic 2/2/0 5/9/3

SM, sensory-motor neuropathy; M, pure-motor neuropathy; MD, motor-dominantneuropathy; CS, corticosteroid; SP, steroid pulse; PP, plasmapheresis; IVIG, intra-venous immunoglobulin.

Table 2 Clinical features in four CIDP patients with aCL

No Age/sex

aCL (U/l)b2-GPI(U/l) Course Therapy

Rankinscore

OtherdiseasesIgG IgM Max Min

1 62/f 43 – – M IVIG 4 1 DM2 53/m 16 – – RR SP 3 1 Lung cancer3 72/m 14 – – RR SP 5 2 MI4 30/f 26 – – M SP 3 0 Alcoholism

M, monophasic; RR, relapsing–remitting; DM, diabetes mellitus; MI, myocardialinfarction; IVIG, intravenous immunoglobulin; SP, steroid pulse.

Table 3 Laboratory, electrophysiologic, and pathologic features of 21 patientswith CIDP

Patients withaCL (n ¼ 4)

Patients withoutaCL (n ¼ 17)

ANA (range) 1520 � 2410(160–5120)

36 � 54(0–160)

Tibial nerveDistal latency (ms) 8.1 � 2.9 5.4 � 1.6MCV (m/s) 32.3 � 5.8* 38.5 � 8.7Sural nerve SCV (m/s) All patients absent 50.2 � 9.4

(seven patients absent)Myelinated nerve fiber density 3817 � 1094* 5904 � 2494Teased fibers

C + D (%) 38.3 � 11.6 29.4 � 20.4E (%) 7.0 � 6.9 0.8 � 0.8

Data are presented as mean � SD. MCV, motor conduction velocity; SCV, sensoryconduction velocity.* P statistically significant compared with patients without aCL.

Nakajima et al.

260

compared with aCL) patients (Table 3). Withregard to sensory nerve findings, the absence ofsural sensory nerve action potential was noted inseven of the 14 aCL) patients. On the contrary, allfour aCL+ patients showed absence of suralsensory nerve action potential. When teased fiberswere classified according to the criteria of Dyck etal. (19), category C, D, and E abnormalities weremore frequent in aCL+ patients (Table 3). Myeli-nated nerve fiber densities were significantly lowerthan those in aCL) patients. These findings suggestthat aCL+ patients may have a more severneuropathy, due to the findings of lower nerveconduction velocities (NCVs) and more severe lossof myelinated fibers in nerve biopsy findings.

Concentrations of aCL before and after treatment

In the four aCL+ CIDP patients, aCL concentra-tions were measured before and after treatment.There is a tendency for aCL levels to decrease afterinitiation of treatment with IVIG or steroid pulsetherapy. The aCL decreased in association withimprovement of neurological disabilities aftertreatment in all four patients (Fig. 1).

Discussion

The present study identified several unique clinico-pathological features of CIDP patients with aCL:

clinically evident sensory-motor polyneuropathy;high ANA titers (over 1:160); marked neuropathybut no vasculitis in the sural nerve; lack ofthromboembolism and collagen disease; and goodresponsiveness to steroid pulse or IVIG treatmentdespite age and concurrent illness. The presentstudy is the first to report the presence of aCL insome patients with CIDP.The aCL are a subset of anti-phospholipid

antibodies frequently detected in sera from patientswith SLE and related autoimmune disorders. Theseautoantibodies are considered responsible for APS,which is characterized by arterial and/or venousthrombosis, recurrent fetal loss, thrombocyto-penia,andavarietyofneurologicsyndromes(15,20).The spectrum of neurologic syndromes associatedwith aCL includes consequences of cerebral throm-boembolism such as transient ischemic attack andcerebral infarction, in addition to neurologic com-plications considered unrelated to thromboembo-lism such as migraine, seizure, chorea, andGBS (16,21). In patients with GBS, Chaleomchan et al. (22)reported 24% prevalence of aCL, and Framptonet al. (23) described a correlation between concen-tration of IgA aCL and GBS severity. They pro-posed that these autoantibodies cross-reacted withphospholipids in nerve, including sphingomyelinand cephalin, initiating damage to the neural struc-tures containing them (24).Patients with various infectious diseases as well

as apparently healthy individuals have been foundto have aCL; these antibodies were first detected byWasserman in patients with syphilis. Recently,aCL have also been found in sera from patientswith non-syphilitic infections such as humanimmunodeficiency virus (25), cytomegalovirus(26), Epstein–Barr virus (EBV) (27), and hepatitisC virus (28). However, clinical manifestations ofAPS do not occur in patients with these infectiousdiseases. The binding specificity of the antibodiesfound in APS is now known to differ from that ofantibodies in infectious conditions. The maindifference appears to be an absolute requirementfor b2-GPI, a plasma protein, in detection of anti-phospholipid antibodies in APS (29, 30). WhileaCL in APS recognize the structure of b2-GPI,aCL in infectious diseases do not (31). Forastieroet al. (31) demonstrated that IgG aCL from APSpatients recognized b2-GPI on irradiated plates inthe absence of phospholipids, while IgG purifiedfrom syphilis patients bound to CL alone. Thus,anti-b2-GPI antibodies were detected in patientswith APS, but not in patients with those infectiousdiseases (32). In this study, anti-b2-GPI antibodieswere not detected in sera from the four aCL+CIDP patients. Our patients with CIDP did not

50

P = 0.04

40

30

20

10

Before

aCL

(U/I)

After

Figure 1. Serum concentration of anti-cardiolipin antibody(aCL) in four CIDP patients with aCL before and after treat-ment. The dashed line represents upper limit of the normalrange.

CIDP with anti-cardiolipin antibodies

261

fulfill the criteria for APS, and lacked any systemicor laboratory features of vasculitis. Thus, aCL inCIDP differs from �autoimmune� aCL detected inAPS patients. These findings suggest that somepatients with CIDP may produce aCL as a result ofmyelin damage or cross-reactions with other anti-myelin antibodies through mechanisms resemblingthose in post-infectious immune responses. Even ifthese aCL represent an epiphenomenon, theyappear to develop a unique subgroup of CIDP.Further detailed neurologic and immunologicstudies should help to resolve these issues.

References

1. Melendez-Vasquez C, Redford J, Choudhary PP, et al.Immunological investigation of chronic inflammatory de-myelinating polyradiculoneuropathy. J Neuroimmunol1997;73:124–34.

2. Prineas JW. Acute idiopathic polyneuritis. An electronmicroscope study. Lab Invest 1972;26:133–47.

3. Schmidt B, Toyka KV, Kiefer R, Full J, Hartung HP,

Pollard J. Inflammatory infiltrates in sural nerve biopsiesin Guillain–Barre syndrome and chronic inflammatorydemyelinating neuropathy. Muscle Nerve 1996;19:474–87.

4. Ho TW, Willison HJ, Nachamkin I, et al. Anti-GD1aantibody is associated with axonal but not demyelinatingforms of Guillain–Barre syndrome. Ann Neurol1999;45:168–73.

5. Kamakura K, Kaida K, Kusunoki S, Miyamoto N, Fukuda J,

Motoyoshi K. Elevation in anti-GQ1b, anti-GT1a, andanti-GT1b IgG antibodies in postinfectious acute ataxicneuropathy with oropharyngeal palsy but without opht-halmoplegia. J Neurol 2000;247:566–7.

6. Kusunoki, S, Chiba A, Kon K, et al. N-acetylgalactosaminylGD1a is a target molecule for serum antibody in Guillain–Barre syndrome. Ann Neurol 1994;35:570–6.

7. Sindern E, Stark E, Haas J, Steck AJ. Serum antibodiesto GM1 and GM3-gangliosides in systemic lupuserythematosus with chronic inflammatory demyelinatingpolyradiculoneuropathy. Acta Neurol Scand 1991;83:399–402.

8. Leger JM, Bouche P, Bolgert F, et al. The spectrum ofpolyneuropathies in patients infected with HIV. J NeurolNeurosurg Psychiatry 1989;52:1369–74.

9. Gorson KC, Ropper AH, Adelman LS, Weinberg DH. Influ-ence of diabetes mellitus on chronic inflammatory demy-elinating polyneuropathy. Muscle Nerve 2000;23:37–43.

10. Simmons Z, Albers JW, Bromberg MB, Feldman EL. Long-term follow-up of patients with chronic inflammatorydemyelinating polyradiculoneuropathy, without and withmonoclonal gammopathy. Brain 1995;118:359–68.

11. Bosboom WM, van-den-Berg LH, Franssen H, et al. Diag-nostic value of sural nerve demyelination in chronicinflammatory demyelinating polyneuropathy. Brain2001;124:2427–38.

12. Kieseier BC, Dalakas MC, Hartung HP. Immune mecha-nisms in chronic inflammatory demyelinating neuropathy.Neurology 2002;59:7S–12S.

13. van-Schaik IN, Vermeulen M, van-Doorn PA, Brand A. Anti-GM1 antibodies in patients with chronic inflammatorydemyelinating polyneuropathy (CIDP) treated with intra-venous immunoglobulin (IVIg). J Neuroimmunol 1994;54:109–15.

14. Chassande B, Leger JM, Younes-Chennoufi AB, et al. Per-ipheral neuropathy associated with IgM monoclonalgammopathy: correlations between M-protein antibodyactivity and clinical/electrophysiological features in 40cases. Muscle Nerve 1998;21:55–62.

15. Cervera R, Piette JC, Font J, et al. Antiphospholipidsyndrome: clinical and immunologic manifestations andpatterns of disease expression in a cohort of 1,000 patients.Arthritis Rheum 2002;46:1019–27.

16. Levine SR, Welch KM. The spectrum of neurologic diseaseassociated with antiphospholipid antibodies. Lupus anti-coagulants and anticardiolipin antibodies. Arch Neurol1987;44:876–83.

17. Ad Hoc Subcommittee of the American Academy ofNeurology AIDS Task Force. Research criteria for diag-nosis of chronic inflammatory demyelinating polyneurop-athy (CIDP). Report from an Ad Hoc Subcommittee ofthe American Academy of Neurology AIDS Task Force.Neurology 1991; 41: 617–8.

18. Mata S, Avanzi G, Lombardo R, Cepparone F, Pinto F, Lolli

F. Anti-GM1, anti-central myelin proteins, and anti-cardiolipin autoantibodies during plasma-exchange inGuillain–Barre syndrome (GBS). J Clin Apheresis1998;13:155–62.

19. Dyck PJ, Giannini C, Lais A. Pathologic alterations ofnerves. In: Dyck PJ, Thomas PK, Griffin JW, Low PA,Poduslo JF, eds. Peripheral neuropathy, 3rd edn.Philadelphia: W. B. Saunders, 1993;514–95.

20. Wilson WA, Gharavi AE, Koike T, et al. Internationalconsensus statement on preliminary classification criteriafor definite antiphospholipid syndrome: report of aninternational workshop. Arthritis Rheum 1999;42:1309–11.

21. Harris EN, Englert H, Derve G, Hughes GR, Gharavi A.

Antiphospholipid antibodies in acute Guillain–Barre syn-drome. Lancet 1983; 1361–2, 8363.

22. Chaleomchan W, Hemachudha T, Sakulramrung R,

Deesomchok U. Anticardiolipin antibodies in patients withrabies vaccination induced neurological complicationsand other neurological diseases. J Neurol Sci 1990;96:143–51.

23. Frampton G, Winer JB, Cameron JS, Hughes RA. SevereGuillain–Barre syndrome: an association with IgA anti-cardiolipin antibody in a series of 92 patients. J Neuro-immunol 1988;19:133–9.

24. Gilburd B, Stein M, Tomer Y, et al. Autoantibodies tophospholipids and brain extract in patients with theGuillain–Barre syndrome: cross-reactive or pathogenic?Autoimmunity 1993;16:23–7.

25. Abuaf N, Laperche S, Rajoely B, et al. Autoantibodies tophospholipids and to the coagulation proteins in AIDS.Thromb Haemost 1997;77:856–61.

26. Labarca JA, Rabaggliati RM, Radrigan FJ, et al. Anti-phospholipid syndrome associated with cytomegalovirusinfection: case report and review. Clin Infect Dis1997;24:197–200.

27. Sorice M, Pittoni V, Griggi T, et al. Specificity of anti-phospholipid antibodies in infectious mononucleosis: arole for anti-cofactor protein antibodies. Clin Exp Immu-nol 2000;120:301–6.

28. Prieto J, Yuste JR, Beloqui O, et al. Anticardiolipin anti-bodies in chronic hepatitis C: implication of hepatitis Cvirus as the cause of the antiphospholipid syndrome.Hepatology 1996;23:199–204.

29. Celli CM, Gharavi AE, Chaimovich H. Opposite b2-glyco-protein I requirement for the binding of infectious andautoimmune antiphospholipid antibodies to cardiolipin

Nakajima et al.

262

liposomes is associated with antibody avidity. BiochimBiophys Acta 1999;1416:225–38.

30. Matsuura E, Igarashi Y, Fujimoto M, et al. Heterogeneityof anticardiolipin antibodies defined by the anticardiolipincofactor. J Immunol 1992;148:3885–91.

31. Forastiero RR, Martinuzzo ME, Kordich LC, Carreras LO.

Reactivity to b2 glycoprotein I clearly differentiates

anticardiolipin antibodies from antiphospholipid syn-drome and syphilis. Thromb Haemost 1996;75:717–20.

32. Visvanathan S, Geczy CL, Harmer JA, McNeil HP. Mono-cyte tissue factor induction by activation of b2-glycopro-tein-I-specific T lymphocytes is associated with thrombosisand fetal loss in patients with antiphospholipid antibodies.J Immunol 2000;165:2258–62.

CIDP with anti-cardiolipin antibodies

263