Antiphospholipid Antibodies and the Antiphospholipid Syndrome

Antiphospholipid Antibodies and the Antiphospholipid Syndrome Clinical Relevance in Neuropsychiatric

Systemic Lupus Erythematosus"

DONATO A L A R C ~ N - S E G O V I A , ~ . ~ BRUNO ESTANOL: GUILLERMO GARC~A-RAMOS: AND

ANTONIO R. VILLAd

bDepartments of Immunology and Rheumatology and cNeurology and Psychiatry

dClinical Epidemiology Unit Instituto Nacional de la Nutricibn Salvador Zubirdn

Mexico City, Mexico

INTRODUCTION

Patients with systemic lupus erythematosus (SLE) may have neuropsychiatric manifestations, and an association between these and the presence of antiphospholipid antibodies (aPL) has been proposed mostly on the bases of small series and case reports.'-7 Among the manifestations found to occur in SLE patients who have aPL are transient ischemic attacks TIA AS),',^^' cerebral infarctions (CI),',',' amaurosis fugax? cerebral venous thromboses; chorea,8-I2 1nigraine,4.~~.'~ seizure^,*^'^ transverse rnyeliti~,~ ischemic optic neuropathy? peripheral neuropathy of several types, psychosis, and diffuse encephalopathy .I5

These mainly anecdotal descriptions have not been substantiated in larger series,I6 which could be due to rarity that prevents them from having enough statistical significance. In a study of 500 SLE patients, we found an association of aPL with arterial occlusions of various kinds, including those causing brain infarcts.I6 However, we did not find an association of aPL with these when considered alone. We have now reevaluated this issue in our cohort of 667 SLE patients." For this we have sought in these patients the association of aPL and/or antiphospholipid syndrome (APS) with the neurological complications of SLE.

amis work was supported by Grants from the Consejo Nacional de Ciencia y Tecnologia and the Vkquez-Shano Fund for research on antiphospholipid cofactor syndromes.

eAddress correspondence to Dr. Donato Alar&n-Segovia, Instituto Nacional de la Nutrici6n Salvador Zubirh, 14000 Mexico D.F., Mexico. Tel: 52 5 573 1127; fax: 52 5 513 2926; e-mail: [email protected].

279

280 ANNALS NEW YORK ACADEMY OF SCIENCES

PATIENTS AND METHODS

Patients

This report is based on our SLE cohort, which includes 667 patients recruited at the Lupus Clinic of the Instituto Nacional de la Nu t r i ch Salvador Zubirh in Mexico City. Patient’s follow-up was integrated both historically (since the first visit to the institute) and prospectively (from 1986 to the cut-off time; December 31 1990).17 Thus, we observed our cohort for 2.039 persodyears. All patients fulfilled four or more classification criteria for SLE of the American College of Rheumatology.18 Neurologic manifestations considered were migraine, seizures, peripheral neuropathy, psychosis, transverse myelitis, chorea, and cerebrovascular disease (transient ischemic attack and/or cerebral infarct). We also considered the category of Sneddon’s syn- drome (cerebrovascular disease plus livedo retic~laris).’~ These neurologic manifesta- tions could be observed at any time during follow-up. As in our previous reports, SLE patients were classified according to levels of IgG and/or IgM aPL (negative, low, and high titers) and number of clinical manifestations related to aPL in order to conform to the APS within SLE (definite, probable, doubtful, and negati~e).”.’~ Clinical manifestations related to aPL were livedo reticularis, thrombocytopenia, recurrent fetal loss, venous thrombosis, hemolytic anemia, arterial occlusion, leg ulcer, pulmonary hypertension, and transverse myelitis. We considered a patient to definitely have APS when there were at least two of these clinical manifestations and high titer aPL (vide infra), and probable APS when there were either two clinical manifestations and low aPL titer or one manifestation and high aPL titer.I7 When a patient presented a cerebral infarct or transverse myelitis, such neurological manifesta- tion was not considered for purposes of classification in the categories of APS.

Methods

Antiphospholipid antibodies were determined in serum by the ELSA method using cardiolipin as an antigen and performed as described by Loizou et al.” and modified by Gharavi et a1.22 and by DelezC et al.” We considered an aPL titer negative when it had a value of less than 2 SD above the mean of 100 healthy subjects in at least one serum sample, low positive aPL when levels were between 22 SD and 15 SD of mean normal values, and high positive when >5 SD of reference values. Antibodies to phospholipid free pz glycoprotein I (p2 GPI) were tested by ELTSA and confirmed by Western blot, as previously described.24

Statistical Analysis

We determined the strength of associations between neurological manifestations and aPL presence (positive vs. negative) and APS (definite vs. negative, definite plus probable vs. negative) by means of odds ratios (OR) calculated as the exponential of coefficients in unconditional logistic regression analyses. To compare differences between anti-P,GPI antibody values with respect to time relationship to thrombosis,

ALARCON-SEGOVIA et UL: aPL: CLINICAL RELEVANCE 281

TABLE 1. Neurological Manifestations in 667 SLE Patients

Manifestation No. Percent

Migraine Seizures Cerebrovascular disease

Transient ischemic attack Cerebral infarct

Peripheral neuropathy Psychosis Sneddon's syndrome Transverse myelitis Chorea

93 59 57 39 24 53 51 26 6 4

13.9 8.8 8.5 5.8 3.6 7.9 7.6 3.9 0.9 0.6

Anv neuroloeical manifestation 238 35.7

we carried out Wilcoxon matched-pairs tests. SPSS/F'C2' was used to perform the mentioned analyses.

RESULTS

Among the 667 SLE patients, we found 101 with definite APS and 143 with probable APS. There were 255 patients who had neither positive aPL nor any of the clinical manifestations that are associated with these antibodies and were thus considered not to have APS. The remaining 168 patients were considered doubtful, either because they had negative aPL but two or more clinical manifestations, low aPL titers and one clinical manifestation, or no clinical manifestations despite having high aPL titers. These were not analyzed further.

The neurological manifestations found in the 667 SLE patients are shown in TABLE 1. In this TABLE we also present the association of cerebrovascular disease with livedo reticularis (Sneddon's ~yndrome), '~ because livedo reticularis is a manifes- tation of APS.l6 Neurological manifestations were present in 35.7% of SLE patients. The number of neurological manifestations present in individual SLE patients are presented in TABLE 2. Ten percent of patients had more than one. When we culled patients with TIAs and CI, we found a total of 57 (8.5%) with cerebrovascular disease (CVD).

There were 444 SLE patients in our lupus cohort who had positive aPL at any level at any one time and 223 who were persistently negative. There was no association between any of the neurological manifestations and aPL positivity (TABLE 3). Surpris- ingly, Sneddon's syndrome was not found to associate with aPL. When we repeated the analysis, considering only the 275 patients with high titer aPL, the lack of association between neurological complications and aPL was also apparent (TABLE 4).

We then tested for the possible association between neurological manifestations and definite APS. Because vascular occlusive disease and transverse myelitis are included in the definition of APS, we excluded patients in which their presence was

282 ANNALS NEW YORK ACADEMY OF SCIENCES

TABLE 2. Culling of Neurological Manifestations“ in 667 SLE Patients

No. Percent

None One Two Three Four Five

429 170 55 10 2 1

64.3 25.5 8.2 1.5 0.3 0.1

At least one 23 8 35.7

a These include migraine, seizures, peripheral neuropathy, psychosis, cerebrovascular disease, transverse myelitis, and chorea.

TABLE 3. Associated Risks of Neurological Manifestations according to Presence of Antiphospholipid (IgG and/or IgM) Antibodies in SLE Patients

Antiphospholipid Antibodies

Positive“ Negative (n = 444) (n = 223)

Neurological Manifestation No. Percent No. Percent ORh 95% CI p value

Migraine 61 13.7 Chorea 4 0.9 Peripheral neuropathy 36 8.1

Seizures 37 8.3 Psychosis 36 8.1 Cerebrovascular disease 41 9.2

Transient ischemic attack 26 5.9 Cerebral infarct 18 4.1

Transverse myelitis 5 1.1

Sneddon’s syndrome 20 4.5

32 0

17 I

22 15 16 13 6 6

14.3 0.93 0 -

7.6 1.06 0.4 1.03 9.9 0.85 6.7 1.21 7.2 1.16 5.8 0.98 2.7 1.10 2.7 1.65

0.59- 1.49 0.77 - 0.31

0.58-1.93 0.86 0.09- 11.37 0.98 0.49- 1.48 0.56 0.65-2.27 0.55 0.63-2.15 0.63 0.49- I .96 0.96 0.41 -2.93 0.85 0.65-4.19 0.29

Any neurological manifestation 163 36.7 75 33.6 1.09 0.77- 1.53 0.63

“ Positive at any level (cutoff mean ? 2 SD of control’s levels). Odds ratio.

essential for the definition of the syndrome. Notwithstanding, we found a significant association between CVD and any neurological manifestation with definite APS (TABLE 5). The latter was related to the fact that all neurological manifestations occurred more frequently in patients with definite APS, but only CVD reached significance.

Seeking to increase our statistical power, we culled patients with definite and probable APS, which together amounted to 236 patients. In so doing, however, we lost significance even for the “any neurological manifestation” category (TABLE 6).

ALARCON-SEGOVIA et UZ. : aPL: CLINICAL RELEVANCE 283

TABLE 4. Associated Risks of Neurological Manifestations according to Presence of High-Level Antiphospholipid (IgG andor IgM) Antibodies in SLE Patients

Antiphospholipid Antibodies

Positive" Negative (n = 275) (n = 223)

Neurological Manifestation No. Percent No. Percent ORh 95% CI p value

Migraine 43 15.6 32 14.3 1.07 Chorea 2 0.7 0 0 - Peripheral neuropathy 25 9.1 17 7.6 1.19 Transverse myelitis 3 1.1 1 0.4 0.83 Seizures 22 8.0 22 9.9 0.81 Psychosis 22 8.0 15 6.7 1.17 Cerebrovascular disease 24 8.7 16 7.2 1.04

Transient ischemic attack 15 5.5 13 5.8 0.89 Cerebral infarct 11 4.0 6 2.7 0.96

Sneddon's syndrome 11 4.0 6 2.7 1.40

0.65- 1.77

0.62-2.27 0.05- 13.32 0.44- 1.51 0.59-2.34 0.54-2.46 0.41 - 1.93 0.32-2.89 0.50 - 3.90

0.78 0.50 0.60 0.89 0.5 1 0.65 0.91 0.76 0.94 0.52

Any neurological manifestation 105 38.2 75 33.6 1.15 0.79-1.67 0.46

' Positive at high level (cutoff mean ? 5 SD of control's levels). Odds ratio.

TABLE 5. Associated Risks of Neurological Manifestations with Definite Antiphospholipid Syndrome in Univariate Logistic Regression

Antiphospholipid Syndrome

Definite" Negative (n = 97) (n = 255)

Neurological Manifestation No. Percent No. Percent ORb 95% CI p value

Migraine Chorea Peripheral neuropathy Transverse myelitis Seizures Psychosis Cerebrovascular disease

Transient ischemic attack Cerebral infarct

Any neurological manifestation

22 2

13 1

11 10 12 8 4

50

22.7 38 14.9 1.68 2.1 1 0.4 5.35

13.4 18 7.1 2.04 1.0 0 0 -

11.3 20 7.8 1.50 10.3 17 6.7 1.61 12.4 15 5.9 2.26 8.2 13 5.1 1.67 4.1 4 1.6 0.84

51.5 83 32.5 2.21

0.93-3.01 0.48-59.66 0.96-4.34

- 0.69-3.27 0.71 -3.65 1.02-5.02 0.67-4.17 0.52- 1.36

1.37-3.55

0.09 0.17 0.07 0.61 0.30 0.26 0.05 0.27 0.47

0.001

" Four patients were excluded because their neurological manifestations contributed to the definition of antiphospholipid syndrome (three with cerebral infarct, and one with transverse myelitis).

Odds ratio.

284 ANNALS NEW YORK ACADEMY OF SCIENCES

TABLE 6. Associated Risks of Neurological Manifestations with Definite Plus Probable Antiphospholipid Syndrome in Univariate Logistic Regression

Antiphospholipid Syndrome Definite + Probable Negative (n = 236)" (n = 255)

Neurological Manifestation No. Percent No. Percent ORb 95% C1 p value

Migraine 30 12.7 38 14.9 0.83 Chorea 3 1.3 1 0.4 3.27 Peripheral neuropathy 21 8.9 18 7.1 1.29

Seizures 24 10.2 20 7.8 1.33 Psychosis 21 8.9 17 6.7 1.37 Cerebrovascular disease 24 10.2 15 5.9 1.81

Transient ischemic attack 15 6.4 13 5.1 1.26 Cerebral infarct I 1 4.7 4 1.6 3.05

Transverse myelitis 1 0.4 0 0 -

0.50-1.39 0.34-31.66 0.67-2.48

0.71 -2.48 0.70-2.66 0.93 -3.54 0.59-2.72 0.96-9.70

0.48 0.31 0.45 0.48 0.37 0.36 0.08 0.55 0.06

Any neurological manifestation 89 37.7 83 32.5 1.26 0.87- 1.82 0.23

a Eight patients were excluded because their neurological manifestations contributed to the definition of antiphospholipid syndrome (five with cerebral infarct, and three with transverse myelitis).

Odds ratio.

TABLE 7. Associated Risk of Antiphospholipid Syndrome (definite vs. negative) with Neurological Manifestations when Controlled by Clinical Manifestations of SLE in Multivariate Logistic Regression"

Variable OR 95% CI p value

Neurological manifestations 1.78 1.06-2.99 0.03 Leukopenia 2.18 1.29-3.70 0.004 Vasculitis 2.14 1.28 -3.57 0.004 Nephropath y 1.89 1.14-3.13 0.01 Alopecia 0.46 0.23 -0.92 0.03

"Model's p-value < 0.001.

In testing models of multivariate logistic regression for associated risks of anti- phospholipid syndrome (definite vs. negative) with neurological manifestations when controlled by clinical manifestations of SLE, we found a positive association of neurological manifestations with leukopenia, vasculitis, and nephropathy, and a nega- tive association with alopecia. This model had a p value of less than 0 . 0 1 (TABLE 7).

In trying to understand the reason for an association between definite APS with a lack of association with aPL, we tested for anti-& GPI antibodies in eight SLE

ALARCON-SEGOVIA et al.: aPL: CLINICAL RELEVANCE 285

TABLE 8. Anti-P,-glycoprotein I Antibodies in Eight SLE Patients with Intracranial Thrombosesa

Anti-P,-glycoprotein I antibodies (OD)”

(time relationship thrombosis) Patient Site of Thromboses Before During After

Venous 1 2

3 4 5 6 7 8

Arterial

Superior longitudinal sinus Superior longitudinal sinus

Temporal-parietal infarction Temporal lobe infarction ParietaVtemporal lobe infarction Anterior brain artery Parietal lobe infarction Internal capsule infarct

0.519 0.432

0.685 0.225 0.857 0.475 0.932 0.418

0.203 0.098

0.138 0.069 0.278 0.069 0.216 0.315

0.297 0.278

0.418 0.323 0.739 0.423 0.41 3 -

Mean Median SD

0.568 0.173 0.413 0.497 0.171 0.413 0.239 0.094 0.156

a Statistical significance by Wilcoxon matched-pairs test: before vs. during, p = 0.01. during vs. after, p = 0.02.

patients with intracranial thrombosis of whom we had sera during the thrombotic episode as well as some time before (8 patients) and after (7 patients). The eight patients were unselected except for their serum availability. However, all eight had positive anti-P2 GPI and, more remarkable, there was a significant fall in the antibody levels at the time of thrombosis, as compared to both previous or subsequent sera (TABLE 8).

DISCUSSION

None of the neurological manifestations present in our SLE patients associated with aPL, and these were as frequent in patients with aPL as in those without. That APS would associate with culled neurological manifestations, as well as with CVD, would seem paradoxical inasmuch as high-titer aPL is one of the criteria for the ~yndr0me.l~

It has recently been shown that aPL detected in conventional ELISA assays are actually reactive with a protein cofactor, P2 GPI, present in the newborn calf serum used in the assay.26 The antibody either recognizes a neoepitope in this protein, which is uncovered by its binding to the anionic or its low avidity permits binding only to a higher density of P2 GPI, which the presence of anionic phospholipid permits.28 Some patients have antibodies that react with p2 GPI in the absence of

286 ANNALS NEW YORK ACADEMY OF SCIENCES

ph0spholipid,2~ and presence of these antibodies in SLE patients associates more strongly with the clinical manifestations of APS than do aPL.24 There are SLE patients with the clinical manifestations of APS but persistently negative aPL in conventional assays who have been found to have persistently positive anti+$ GPI.30 This we have considered a variant of antiphospholipidkofactor syndromes?' which may also have a primary variant.32

We did a preliminary search for anti-& GPI in patients with intracranial thrombo- ses and found them to be uniformly positive. Moreover, there was a remarkable fall in antibody levels at the time of thrombosis in the eight patients whom we tested, both before and during the episode. The subsequent rise in levels after the thrombotic event could indicate an otherwise relatively constant antibody positivity.

We could interpret our findings as indicative of an association between CVD in lupus patients with anti+, GPI rather than with aPL. The association with APS where aPL is a requirement could mean that these two autoantibodies tend to ride together. Indeed, it is quite likely that both of these are directed to the same or related epitopes in p2 GPI and differ only in their immunochemical characteristics. In this respect they would be similar to anti-DNA antibodies, which are more likely to associate with disease activity and renal damage when of high avidity.

It is clear that considerably more study is needed in order to determine if anti- p2 GPI are more clearly associated with the neurological manifestations of SLE than are aPL.

ACKNOWLEDGMENTS

The help of Javier Cabiedes and Antonio R. Cabral is gratefully recognized.

1.

2.

3.

4.

5 .

6.

7.

8.

9.

REFERENCES

HARRIS, E. N. 1984. Cerebral infarction in systemic lupus: Association with anticardiolipin antibodies. Clin. Exp. Rheumatol. 2: 47-51.

ASHERSON, R. A. et al. 1989. Cerebrovascular disease and antiphospholipid antibodies in systemic lupus erythematosus, lupus-like disease, and the primary antiphospholipid antibody syndrome. Am. J. Med. 8 6 391-399.

LEVINB, S. R. et al. 1987. Cerebral venous thrombosis with lupus anticoagulants. Report of two cases. Stroke 18: 801-804.

LEVINE, S. R. et al. 1987. Migraine and the lupus anticoagulant. Case reports and review of the literature. Cephalalgia 7: 93 -99.

LAVALLE, C. et al. 1990. Transverse myelitis: A manifestation of systemic lupus erythema- tows strongly associated with antiphospholipid antibodies. J. Rheumatol. 17: 34-37.

JABS, D. A. et al. 1986. Optic neuropathy in systemic lupus erythematosus. Arch. Ophthal- mol. 104: 564568.

ASHERSON, R. A. et al. 1987. Recurrent stroke and multi-infarct dementia in systemic lupus erythematosus associated with antiphospholipid antibodies. Ann. Rheum. Dis.

HODGES, J. R. 1987. Chorea and the lupus anticoagulant. J. Neurol. Neurosurg. Psychiatry

LUBBE, W. F. & E. B. WALKER. 1988. Chorea gravidarum associated with circulating lupus anticoagulant. Successful outcome of pregnancy with prednisone and aspirin therapy. Br. J. Obstet. Gynaecol. 90: 487-490.

4 6 605-611.

50: 368-369.

ALARCON-SEGOVIA et UL : aPL: CLINICAL RELEVANCE 287

10.

11.

12.

13. 14. 15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

ICHIKAWA, K. e l al. 1980. Chorea gravidarum: Report of a fetal case with neuropathological observations. Arch. Neurol. 37: 429-432.

VLACHOYIANNOPOULOS, P. G., G. DIMOU & A. SIAMOPOULOU-MAVDRIDOU. 1991. Chorea as a manifestation of the antiphospholipid syndrome in childhood. Clin. Exp. Rheurnatol.

ASHERSON, R. A. er al. 1987. Chorea in systemic lupus erythematosus and “lupus like” disease: Association with antiphospholipid antibodies. Semin. Arthritis Rheum. 16:

HERING, R. et al. 1991. Antiphospholipid antibodies in migraine. Cephalalgia. 11: 19-21. SHUAIB, A. et al. 1987. Migraine and antiphospholipid antibodies. Headache 29: 42-43. LOVE, P. E. & S. A. SANTORO. 1990. Antiphospholipid antibodies: Anticardiolipin and

the lupus anticoagulant in systemic lupus (SLE) erythematosus and non-SLE disorders. Ann. Intern. Med. 112: 682-698.

ALARCON-SEGOVIA, D. er al. 1989. Antiphospholipid antibodies and the antiphospholipid syndrome in systemic lupus erythematosus. A prospective analysis of 500 consecutive patients. Medicine 68: 353-365.

ALARCON-SEGOVIA, D. et al. 1992. Preliminary classification criteria for the antiphospho- lipid syndrome within systemic lupus erythematosus. Semin. Arthritis Rheum. 21:

TAN, E. M. et al. 1982. The 1982 revised criteria for the classification of systemic lupus

SNEDDON, I. B. 1965. Cerebrovascular lesions and livedo reticularis. Br. J. Dermatol. 77:

PEREZ-VAZQUEZ, M. E. et al. 1993. Influence of disease duration, continued follow-up and further antiphospholipid testing on the frequency and classification category of antiphospholipid syndrome in a cohort of patients with systemic lupus erythematosus. J. Rheumatol. 2 0 437-442.

LOIZOU, S. 1985. Measurement of anticardiolipin antibodies by an enzyme-linked immuno- sorbent assay (ELISA): Standardization and quantization of results. Clin. Exp. Rheuma-

GHARAVI, A. E. et al. 1987. Anticardiolipin antibodies: Isotype distribution and phospho- lipid specificity. Ann. Rheum. Dis. 46: 1-6.

DELEZE, M., C. V. ORIA & D. ALARCON-SEGOVIA. 1988. Occurrence of both hemolytic anemia and thrombocytopenic purpura (Evans’ syndrome) in systemic lupus erythema- tosus. Relationship to antiphospholipid antibodies. J. Rheumatol. 15: 61 1 -615.

CABIEDES, J., A. R. CABRAL & D. ALARCON-SEGOVIA. 1995, Clinical manifestations of the antiphospholipid syndrome in systemic lupus erythematosus associate more strongly with anti-P2-glycoprotein- 1 than with antiphospholipid antibodies. J. Rheumatol. 22:

NORUSIS, M. J. 1992. SPSS/PC+. Base System User’s Guide. Version 5.0. SPSS Inc. Chicago, III.

MCNEIL, H. P. et al. 1990. Anti-phospholipid antibodies are directed against a complex antigen that includes a lipid-binding inhibitor of coagulation: P,-Glycoprotein I (apolipo- protein H). Proc. Natl. Acad. Sci. USA 87: 4120-4124.

MAATSURA, E. et al. 1994. Anticardiolipin antibodies recognize P,-Glycoprotein I structure altered by interacting with an oxygen modified solid phase surface. J. Exp. Med. 179:

ROUBEY, R. A. S. er al. 1995. “Anticardiolipin” autoantibodies recognize P,-Glycoprotein I in the absence of phospholipid. Importance of Ag density and bivalent binding. J. Immunol. 154: 954-960.

VIARD, J. P., 2. AMOURA & J. F. BACH. 1992. Association of anti-p,-glycoprotein I antibodies with lupus-type circulating anticoagulant and thrombosis in systemic lupus erythematosus. Am. J. Med. 93: 181-186.

9: 303-305.

253-259.

275-286.

erythernatosus. Arthritis & Rheum. 21: 275-286.

180-185.

tol. 62: 738-745.

1899-1906.

457-462.

288 ANNALS NEW YORK ACADEMY OF SCIENCES

30. ALARCON-SEGOVIA, D. et al. 1997. The antiphospholipidcofactor syndromes. 11. A variant in systemic lupus erythematosus patients with antibodies to P,-glycoprotein-I but no antibodies detectable in standard antiphospholipid assays. J. Rheumatol. In press.

ALARCON-SEGOVIA, D. & A. R. CABRAL. 1996. (Editorial) The antiphospholipidcofactor syndromes. J. Rheumatol. 23: 1319-1322.

CABRAL, A. R. et al. 1996. The antiphospholipidcofactor syndromes. I. A primary variant with antibodies to P2-glycoprotein-I but no antibodies detectable in standard antiphos- pholipid assays. Am. J. Med. 101: 472-481.

31.

32.