Antiphospholipid antibodies are a group of autoantibodies that bind to negatively charged phospholipids that have been associated with thrombotic events that could lead to pregnancy loss.1 Clinical features, such as venous or arterial thrombosis, recurrent pregnancy loss, a single fetal loss, or early onset of severe pregnancy hypertension, in conjunction with a positive laboratory finding (positive immunoglobulin [Ig] G or IgM cardiolipin antibodies or a positive lupus anticoagulant) satisfy the criteria for diagnosis of the antiphospholipid antibody syndrome.2 The pathophysiologies, clinical manifestations, and management guidelines related to antiphospholipid antibodies have been reviewed.3
In general, pregnancy exerts its effects on the plasma concentration and activities of several proteins involved in blood coagulation. These changes in conjunction with circulating antiphospholipid antibodies are thought to induce thrombosis by binding to phospholipids on the surface of platelets and the vascular endothelium. This binding complex is characterized by decreased prostacyclin production by endothelial cells, increased thromboxane production by platelets, and decreased protein C activation, resulting in vasoconstriction.4,5
Management of antiphospholipid antibodies during pregnancy has included the use of heparin, aspirin alone, prednisone, and intravenous γ-globulin; however, the combination of subcutaneous heparin and low-dose aspirin has provided the highest success rates relative to other treatments with a low occurrence of side effects.6 Two prospective, controlled trials have shown successful pregnancy rates of 70–75% with unfractionated heparin in combination with aspirin, versus less than 45% for patients who received aspirin alone.7,8 The rationale for heparin use has been that the anticoagulant activity overrides the thrombotic actions caused by antiphospholipid antibody binding to phospholipids, β2-glycoprotein 1, or other cross-reactive substances.4,5,9 The addition of unfractionated heparin to antiphospholipid antibody–positive sera from women with recurrent pregnancy loss caused a dose-dependent decrease in IgG binding to cardiolipin and phosphatidylserine in an enzyme-linked immunosorbent assay (ELISA).10,11 Increased unfractionated heparin taken during pregnancy correlated inversely with the fall in serum antiphospholipid antibody titers in vivo.10
Heparin is a polymer of acidic, sulfated disaccharides, derived from porcine or bovine mucosa. The length of the polysaccharide chain determines the properties of the molecule—shorter chains are low molecular weight heparins and longer chains are high molecular weight (or unfractionated) heparins.12 The length of the chain also determines the properties of the molecule. Heparin binds to and potentiates the action of antithrombin III, inducing a conformational change. Unfractionated heparin binds thrombin and antithrombin III simultaneously, facilitating the inactivation of thrombin. The conformational change in antithrombin III induced by heparin allows the molecule to bind to, and inactivate, factors involved in the clotting cascade. Low molecular weight heparins are more selective inhibitors of factors IXa and Xa than unfractionated heparin.12
The success of unfractionated heparin with pregnancy outcomes in women with antiphospholipid antibody syndrome has stimulated interest in low molecular weight heparin because of several clinical and practical advantages. Low molecular weight heparin has been advocated because of its presumed mechanism of action against thrombosis and absence of frequent monitoring during pregnancy.12 In addition, it has a lower incidence of bone loss and the potential for once-a-day administration. Recent reviews that described the use of low molecular weight heparin therapy in pregnant patients with other clinical conditions have stressed the need for more information on low molecular weight heparin.6,12
Four main possibilities exist to explain the therapeutic effect of heparin toward decreasing recurrent pregnancy loss. Heparin could 1) immunomodulate cell-mediated or humoral events to prevent production or alter the action of antiphospholipid antibodies, 2) produce an antithrombotic effect independent of antiphospholipid antibodies that overrides antiphospholipid antibody action, 3) block the action of antiphospholipid antibodies directly or indirectly, or 4) facilitate the elimination of antiphospholipid antibodies.3–5,9
It is unknown whether the inhibitory effect of unfractionated heparin on in vitro antiphospholipid antibody binding exists with low molecular weight heparin. The low molecular weight heparins are not able to bind both antithrombin III and thrombin simultaneously, which is necessary for the inactivation of thrombin and is the primary mechanism of action of unfractionated heparin.12 Because of this known difference in binding activity, we desired to compare the inhibitory activities on a per unit basis of unfractionated heparin and low molecular weight heparin with regard to in vitro binding of antiphospholipid antibodies from the sera of patients with recurrent pregnancy loss. We report the inhibition of IgG cardiolipin and IgG phosphatidylserine antibody binding by low molecular weight and unfractionated heparin in a dose-response fashion measured by an in vitro ELISA.
MATERIALS AND METHODS
Sera from women were eligible for inclusion in this study based on a history of three or more spontaneous losses by the same partner and positive levels of IgG cardiolipin and/or IgG phosphatidylserine antibodies repeated on two occasions. Demographic data for the women who donated sera for this study are included in Table 1. For this study, only serum samples with 30 or more IgG phospholipid units were selected so that inhibition could be evaluated. All samples were obtained from nonpregnant women who were not on heparin. The sera were obtained from women who were patients at the Recurrent Pregnancy Loss Center at the University of Tennessee. This study was approved by the University of Tennessee Institutional Review Board.
All serum samples were evaluated for IgG antibodies against cardiolipin and phosphatidylserine utilizing the ELISA method as described by Harris.13 Briefly, individual 96-well microtiter plates (Immulon-2; Dynatech Labs, Chantilly, VA) were coated with 30 μL of purified phospholipids (Sigma Chemical Co., St. Louis, MO) at a concentration of 45 μg/mL (cardiolipin) in ethanol or 50 μg/mL (phosphatidylserine) in methanol. The plates were air dried overnight at 4C and then blocked with 200 μL of 10% fetal calf serum (GIBCO, Grand Island, NY) in phosphate-buffered saline (GIBCO), washed, and incubated at 37C for 2 hours with 50 μL of patients' sera diluted 1:50 in 10% fetal calf serum in phosphate-buffered saline. Each unknown sample was run in duplicate. The plates were then washed to remove unbound antibody and proteins, and a secondary antibody, alkaline-phosphatase–conjugated antihuman IgG (Caltag Labs, San Francisco, CA), was added to the plate. After incubation and washing, p-nitrophenyl phosphate substrate (Sigma 104) was added and used to indirectly measure the level of antiphospholipid antibodies in a patient's serum. The optical density of the samples, caused by the cleavage of the substrate by the enzyme, was determined at a dual wavelength of 405/550 nm by a Bio-Tek Microplate Reader (model EL 340; Bio-Tek Instruments, Winooski, VT) and was used to quantify the amount of antiphospholipid antibodies in the sera.
Every assay plate also included a known highly positive anticardiolipin sample (more than 100 IgG phospholipid units). Plates were incubated until the highly positive wells achieved an optical density of greater than 1.0; typically, this required an incubation of 20–30 minutes. Referenced standard sets for cardiolipin (Louisville APL Diagnostics, Louisville, KY) and known negative sera were used on every plate. All results were defined in phospholipid units for IgG as follows: less than 10 units, negative; 10–19 units, borderline; 20–80 units, positive; and more than 80 units, highly positive. Phosphatidylserine values were interpreted based on the multiples of the median method as described previously.14
Briefly, phospholipid units for IgG were calculated for each serum sample, and the median value was determined from the nongaussian distribution. The cutoff value in phospholipid units of phosphatidylserine was determined by using the 99th percentile of the normal population, approximately 3.0 times the median value. All values reported as positive were the means of duplicated determinations with background absorbance obtained from wells prepared without the coating phospholipid subtracted. Any values with a standard error greater than 10% were discarded and reassayed. Interassay variation was less than 8%, and intraassay variation was less than 6%.
The standard antiphospholipid antibody ELISA was modified to observe the inhibitory effect of heparin on antiphospholipid antibodies. Various dilutions of unfractionated heparin sodium (Elkins-Sinn Inc., Cherry Hill, NJ) and low molecular weight heparin (Sigma) were prepared to contain 0, 16, 32, 64, 128, and 256 units in phosphate-buffered saline. Plates were blocked and washed as described above. Patients' diluted sera and 100 μL of the heparin solutions were added to the plates simultaneously. The assay was then completed as described above. Preliminary studies indicated this range of heparin (0–256 units) to be appropriate to evaluate the inhibition of antiphospholipid antibody binding (data not shown). All values had background absorbances subtracted.
Each individual sample of patient sera was fractionated over unfractionated and low molecular weight heparin affinity columns with a bed volume of 1.0 mL.10 Columns containing unfractionated heparin that was cross-linked to 4% beaded agarose were purchased commercially (Sigma HEP-II-5). Columns containing low molecular weight heparin were prepared by cross-linking low molecular weight heparin (Sigma H-3400) to 4% cyanogen bromide–activated agarose beads (Sigma C9210). Columns were equilibrated with at least three volumes of 0.01-mmol/L tris-hydrochloride buffer (pH 7.5). Serum samples (50 μL) were diluted with 250 μL of equilibrating buffer, loaded onto individual columns, and run in equilibrating buffer. The columns were then eluted with 0.01-mol/L tris-hydrochloride buffer (pH 7.5) containing 1.5-mol/L sodium chloride. Fractions of 1.0 mL were collected and analyzed by ulttraviolet spectrophotometry at 280 nm. The fractions containing the protein that passed through the column (“passed”) and the fractions that were absorbed to heparin and then eluted (“eluted”) were pooled separately and assayed for antiphospholipid antibody activity by a standard ELISA. Additional controls in the ELISA of these samples consisted of nonfractionated serum, equilibration buffer, and elution buffer.
Statistical analyses, using the Fisher two-tailed exact test, were performed on patient demographics and the 50% inhibition analysis (Tables 1 and 2). For some data analysis, patients' sera were subdivided into two groups based on the levels of antiphospholipid antibody activity detected. Sera samples were grouped as “moderately positive” if they contained 30–79 IgG phospholipid units, whereas samples with 80–200 IgG phospholipid units were grouped as “highly positive.” The repeated-measures analysis of variance was performed on the unfractionated and low molecular weight heparin doseresponse inhibition curves (Figures 1 and 2). Using the analysis of variance, a comparison of measured phospholipid units was made between unfractionated and low molecular weight heparin at each point on the dose-response curve. In addition, a comparison of the measured phospholipid units with no heparin added versus the various concentrations of unfractionated and low molecular weight heparin (16–256 units) was made.
There were no differences in patient demographics between the women with antibodies to cardiolipin and those with antibodies to phosphatidylserine in their sera using the Fisher exact test (Table 1).
Increasing amounts of unfractionated and low molecular weight heparin (0, 16, 32, 64, 128, and 256 IU) inhibited the binding of IgG cardiolipin and IgG phosphatidylserine in sera obtained from women with recurrent pregnancy loss (Figure 1). There were no significant differences observed between heparin derivatives at each individual concentration evaluated; however, in women with moderate levels of IgG cardiolipin and IgG phos-phatidylserine (30–79 IgG phospholipid units) a dose-dependent decrease in antiphospholipid antibody binding was observed. With the addition of 256 IU of unfractionated heparin or low molecular weight heparin there was maximal inhibition of 76–89%. Significant inhibition of IgG cardiolipin (P < .001) and IgG phos-phatidylserine (P < .05) activity was observed with as little as 32 IU of low molecular weight heparin and in the presence of 64 IU of unfractionated heparin (P < .001 and P < .05, respectively).
In women with high levels of antiphospholipid antibodies (80–200 IgG phospholipid units), 82–89% of the IgG cardiolipin and IgG phosphatidylserine activity was maximally reduced with 256 IU of heparin (Figure 2). A significant decrease in IgG cardiolipin binding activity (Figure 2A) was achieved with the addition of 128 IU of each heparin analogue (P < .01). Similarly, in sera with high levels of IgG phosphatidylserine only 64 IU of each heparin analogue were required for significant reduction in antiphospholipid antibody in vitro binding (P < .01).
The amount of unfractionated heparin and low molecular weight heparin that inhibited antiphospholipid antibody binding by half was ascertained (Table 2). There were 14 serum samples each for IgG cardiolipin and IgG phosphatidylserine that were assayed and grouped based on “highly” or “moderately” positive levels of antiphospholipid antibodies. On a per unit basis, moderate IgG cardiolipin binding (44.1 ± 2.7 IgG phospholipid units) was reduced by 50% with 89.6 ± 30.1 IU of low molecular weight heparin, compared with 144.6 ± 39.8 IU of unfractionated heparin (P = .29). Similarly, high IgG phosphatidylserine activity (124.1 ± 17.1 IgG phospholipid units) was decreased, with 90.6 ± 16.9 IU of low molecular weight heparin versus 121.1 ± 34.0 IU of unfractionated heparin (P = .18). Moreover, low molecular weight heparin reduced antiphospholipid antibody binding as well as unfractionated heparin among subjects with moderate IgG phosphatidylserine activity (47.6 ± 6.0 IgG phospholipid units) with the addition of 91.1 ± 38.3 IU of low molecular weight heparin and 88.± 37.5 IU of unfractionated heparin (P = .96). High IgG cardiolipin activity (168.6 ± 20.4 IgG phospholipid units) was inhibited by 50% with the addition of similar amounts of low molecular weight heparin (117.4 ± 17.5 IU) and unfractionated heparin (112.9 ± 4.1 IU) (P = .81).
Serum samples from all women with antiphospholipid antibodies were individually passed over low molecular weight and unfractionated heparin agarose affinity columns. Fractions were divided into two pools: 1) the unbound material that “passed” through the heparin affinity column (fractions 3–5) and 2) the bound material that was absorbed to heparin that was “eluted” with a high salt buffer (fractions 6–8). A typical chromatography profile is shown in Figure 3.
Figure 4 illustrates the amount of IgG cardiolipin and IgG phosphatidylserine activity recovered in fractions from both heparin columns. Elution of antiphospholipid antibodies with high salt solutions from unfractionated and low molecular weight heparin columns recovered 72% and 66% of IgG cardiolipin activity, respectively. By comparison, 46% and 54% of the original IgG phosphatidylserine activity were eluted from the unfractionated and low molecular weight heparin columns. Both the unfractionated and low molecular weight heparin affinity columns bound a greater percentage of the IgG cardiolipin.
This study demonstrates the dose-dependent inhibition of in vitro IgG cardiolipin and IgG phosphatidylserine binding activity in the presence of heparin. In the sera of patients with recurrent pregnancy loss and antiphospholipid antibodies, it was determined that doses of unfractionated or low molecular weight heparin within a range of 32–256 IU were sufficient to significantly decrease IgG cardiolipin and IgG phosphatidylserine binding in an ELISA. These data correlate with previous reports that evaluated the in vitro interactions of unfractionated heparin with antiphospholipid antibody–positive sera.10,11
This study demonstrates that similar amounts of low molecular weight heparin and unfractionated heparin are required to decrease in vitro antiphospholipid antibody binding by 50%. Moreover, the amount of antiphospholipid antibody binding activity inhibited at individual heparin dosages does not significantly differ when the heparin derivatives are compared. This is an important in vitro observation when it is considered that low molecular weight heparin has been reported to exhibit less binding to plasma proteins than unfractionated heparin.15 Generally, unfractionated heparin binds to any positively charged protein because of the negative charge associated with the molecule. Because of less protein binding in plasma, low molecular weight heparin is thought to have a more predictable anticoagulant effect. Low molecular weight heparins are more selective inhibitors of factors IXa and Xa than unfractionated heparins and have less effect on the inhibition of thrombin. This primarily results because their smaller size is not conducive to the concurrent binding of antithrombin III and thrombin.12 These differences do not appear to be significant in vivo during treatment of patients with antiphospholipid antibodies and recurrent pregnancy loss. Clinical investigations have shown that women with antiphospholipid antibodies and recurrent pregnancy loss can expect a live-birth rate of 75% when treated with unfractionated heparin in combination with aspirin.6–8 Several studies suggest that a similar live-birth rate could be achieved with low molecular weight heparin.16,17 Farquharson et al17 recently reported no differences in live-birth rates when comparing women with recurrent pregnancy loss and antiphospholipid antibodies who were treated with low molecular weight heparin and aspirin versus aspirin alone. Although their study was prospective and randomized, it had a number of major flaws that render their conclusions invalid. Many women were randomized after documentation of fetal cardiac activity, almost 25% of the women switched treatment after their random assignment, and the study lacked sufficient power. Moreover, only 11 of 98 women in the study (11%) satisfied criteria for antiphospholipid antibody syndrome as established by the most recent international consensus conference.2
Numerous physiologic mechanisms have been proposed to explain the beneficial effects of heparin in the antiphospholipid antibody syndrome.3–5,12,18 This study suggests that both unfractionated and low molecular weight heparin may directly block the binding of antibodies to phospholipids. In our modified ELISA, we added heparin to the wells at the same time that we added the positive patient sera. However, an indirect mechanism of heparin action is also possible in vivo. Investigators who have followed antiphospholipid antibody levels during pregnancy in women with antiphospholipid antibody syndrome have recorded declining levels of antiphospholipid antibodies.10,18 This observation suggests that heparin may facilitate the clearance of antiphospholipid antibodies in vivo.10,18 In vivo, heparin is metabolized according to a saturable pathway that depends on dosage, degree of plasma protein binding, and molecular weight. The elimination of unfractionated heparin is faster than that of low molecular weight forms, and therefore the duration of action is shorter. If clearance of antiphospholipid antibodies after heparin binding is a contributing mechanism, molecular forms of heparin may be a factor in vivo. Other studies have compared heparin interactions with platelet-endothelial cells and plasma proteins.15,19 These studies indicate that a possible antithrombotic effect of heparin independent of antiphospholipid antibodies overrides the harmful action of antiphospholipid antibodies.20 It is unclear if any single mechanism is responsible for the beneficial effects of heparin in the treatment of antiphospholipid antibody syndrome.
For pregnant women, antiphospholipid antibodies may contribute to serious maternal and fetal complications. Recent reports have increased the awareness of the need for heparin therapy during pregnancy aimed at decreasing the risk of thrombosis and pregnancy loss. Low molecular weight heparin has several potential advantages over unfractionated heparin, including a decreased risk of osteopenia, less laboratory monitoring, once a day dosing, and decreased bleeding when compared with unfractionated heparin.12 Based on this study we determined that on a per unit basis low molecular weight heparin is as effective as unfractionated heparin in inhibiting the binding of antiphospholipid antibodies in vitro. Although these data suggest that low molecular weight heparin and unfractionated heparin have similar binding activities to antiphospholipid antibodies in vitro, they do not address the effectiveness of low molecular weight heparin in the treatment of antiphospholipid antibody syndrome. Ideally, properly designed prospective, randomized controlled trials may correlate with these data in vivo.
1. Lockshin MD. Antiphospholipid antibody. JAMA 1997; 277:1549–51.
2. Wilson WA, Gharavi AE, Piette JC. International classification criteria for antiphospholipid syndrome: Synopsis of a post-conference workshop held at the Ninth International (Tours) aPL Symposium. Lupus 2001;10:457–60.
3. Kutteh WH, Rote NS, Silver R. Antiphospholipid antibodies and reproduction: The antiphospholipid antibody syndrome. Am J Reprod Immunol 1999;41:133–52.
4. Chamley LW, McKay EJ, Pattison NS. Inhibition of heparin/antithrombin III cofactor activity by anticardiolipin antibodies: A mechanism for thrombosis. Thromb Res 1993;71:103–11.
5. Shibata S, Harpel P, Bona C, Filit H. Monoclonal antibodies to heparin sulfate inhibit the formation of thrombinantithrombin III complexes. Clin Immunol Immunopathol 1993;67:264–72.
6. Empson M, Lassere M, Craig JC, Scott JR. Recurrent pregnancy loss with antiphospholipid antibody: A systematic review of therapeutic trials. Obstet Gynecol 2002;99:135–44.
7. Kutteh WH. Antiphospholipid antibody-associated recurrent pregnancy loss: Treatment with heparin and low-dose aspirin is superior to low-dose aspirin alone. Am J Obstet Gynecol 1996;174:1584–9.
8. Rai R, Cohen H, Dave M, Regan L. Randomized, controlled trial of aspirin and aspirin plus heparin in pregnant women with recurrent miscarriage associated with phospholipid antibodies (or antiphospholipid antibodies). BMJ 1997;314:253–7.
9. Tanne D, Triplett DA, Levine SR. Antiphospholipid-protein antibodies and isthmic stroke: Not just cardiolipin anymore. Stroke 1998;29:1755–8.
10. Ermel LD, Marshburn PB, Kutteh WH. Interaction of heparin with antiphospholipid antibodies (APA) from the sera of women with recurrent pregnancy loss (RPL). Am J Reprod Immunol 1995;33:14–20.
11. McIntyre JA, Wagenknecht DR. Interaction of heparin with β2-glycoprotein I and antiphospholipid antibodies in vitro. Thromb Res 1992;68:495–500.
12. Ensom MH, Stephenson MD. Low molecular weight heparins in pregnancy. Pharmacotherapy 1999;19:1013–25.
13. Harris EN. Annotation: Antiphospholipid antibodies. Br J Haematol 1990;74:1–9.
14. Kutteh WH, Wester R, Kutteh CC. Multiples of the median: Alternate methods for reporting antiphospholipid antibodies in women with recurrent pregnancy loss. Obstet Gynecol 1994;84:811–5.
15. Young E, Wells P, Holloway S, Weitz J, Hirsh J. Ex-vivo and in-vitro evidence that low molecular weight heparins exhibit less binding to plasma proteins than unfractionated heparin. Thromb Haemost 1994;71:300–4.
16. Nelson-Piercy C, Letsky EA, deSwiet M. Low molecular weight heparin for obstetric thromboprophylaxis: Experience of sixty-nine pregnancies in sixty-one women at high risk. Am J Obstet Gynecol 1997;176:1062–8.
17. Farquharson RG, Quenby S, Greaves M. Antiphospholipid syndrome in pregnancy: A randomized, controlled trial of treatment. Obstet Gynecol 2002;100:408–13.
18. Masamoto H, Toma T, Sakumoto K, Kanazawa K. Clearance of antiphospholipid antibodies in pregnancies treated with heparin. Obstet Gynecol 2001;97:394–8.
19. Young E, Venner T, Ribau J, Shaughnessy S, Hirsh J, Podor TJ. The binding of unfractionated heparin and low molecular weight heparin to thrombin-activated human endothelial cells. Thromb Res 1999;96:373–81.
20. Rand JH, Wu XX, Andree H, Lockwood CJ, Guller S, Scher J, et al. Pregnancy loss in the antiphospholipid antibody syndrome: A possible thrombogenic mechanism. N Engl J Med 1997;337:154–60.