A prospective, controlled multicenter study on the obstetric risks of pregnant women with antiphospholipid antibodies HenkJ. Out, MD,.,b Hein W. Bruinse, MD,. Godelieve C.M.L. Christiaens, MD,. Marja van Vliet,.' Philip G. de Groot, PhD: H. Karel Nieuwenhuis, MD: and Ronald H.W.M. Derksen, MD b C
Utrecht, The Netherlands OBJECTIVES: A prospective, controlled multicenter study was performed to estimate the obstetric risks of anti phospholipid antibodies (the lupus anticoagulant and anticardiolipin antibodies), In addition, the risks of prior thrombosis, obstetric history, systemic lupus erythematosus, and high-dose prednisone treatment were evaluated, STUDY DESIGN: After screening for anti phospholipid antibodies in patients with lupus erythematosus or women with prior fetalloss(es), 59 subsequent pregnancies with and 54 without these antibodies were followed, RESULTS: The presence of the lupus anticoagulant and a history of at least three spontaneous abortions could predict fetal loss (p = 0,032 and 0,001, respectively), In live born infants, a low birth weight could be predicted by the presence of anticardiolipin antibodies (p = 0,034), prior intrauterine fetal death (p = 0,025), and treatment with high-dose prednisone (p = 0,002). No relationships were seen between anti phospholipid antibodies and small-for-gestational-age newborns and pregnancy-induced hypertension or preeclampsia. The disappearance of anti phospholipid antibodies during pregnancy was not correlated with live birth. CONCLUSION: It is concluded that the presence of antiphospholipid antibodies is a risk factor for adverse pregnancy outcome. (AM J OBSTET GVNECOL 1992;167:26-32.)
Key words: Lupus anticoagulant, anticardiolipin antibodies, pregnancy outcome, systemic lupus erythematosus In the last decade anti phospholipid antibodies, notably the lupus anticoagulant and anticardiolipin antibodies, have received much attention. Their presence has been associated with prior fetal loss, thrombosis, and thrombocytopenia in retrospective reports. ' -3 These correlations have been described in patients with systemic lupus erythematosus, lupuslike disease, and fetal loss without underlying disease." 5 In patients with anti phospholipid antibodies, fetal loss is generally accompanied by placental thrombosis. However, the pathogenetic potential of these antibodies has not been demonstrated." There are no large-scale studies evaluating the obstetric risks for pregnant women with anti phospholipid antibodies in a prospective, controlled study. However, on the basis of studies on small series or case reports, From the Department of Obstetrics and Gynecology: the Division of Immunopathology, Department of Internal Medicine,' and the Department of Hematology,' University Hospital Utrecht. Supported by a grant from the Praeventiefonds (No. 28-1511) to HIO. Received for publication July 10, 1991,- revised December 11,1991,accepted December 30,1991. Reprint requests: HI Out, MD, Department of ObstetriC,l and Gynecology, University Hospital Utrecht, P.O. Box 85500, 3508 GA Utrecht, The Netherlands_ 611 /37391
26
it has become current practice to offer some kind of treatment. Successes have been claimed in favor of a therapy that includes high-dose prednisone," 8 but its benefits are also controversial. 9 Other successful treatments reported include subcutaneous heparin or lowdose aspirin. 9 . 10 The need for large prospective, controlled studies in pregnant women with anti phospholipid antibodies is universally recognized. II. 12 In this study we prospectively evaluated the risks of lupus anticoagulant and anticardiolipin antibodies (both level and isotype) on pregnancy outcome in 59 pregnancies complicated by anti phospholipid antibodies compared with 54 control pregnancies. We also estimated the obstetric risks of systemic lupus erythematosus, lupuslike disease, prior adverse pregnancy outcomes, history of thrombosis, and inclusion of high-dose prednisone in the treatment regImen. Material and methods
Patients. From June 1987 women who wished to conceive and who were from risk groups were screened for the presence of anti phospholipid antibodies to identify eligible patients for this study. Patients were obtained from all over the country. Risk groups included patients with systemic lupus erythematosus according
Volume 167 Number I
to American Rheumatism Association-criteria"; patients with lupuslike disease (fewer than four criteria); and patients with at least three unexplained spontaneous abortions before 12 weeks of gestation or at least one intrauterine fetal death after 12 weeks not caused by infection, abruptio placentae, apparent congenital malformations, or chromosomal abnormalities. In all patients with at least three spontaneous abortions hysterosalpingography was performed to exclude uterine malformations. A total of 264 patients were screened, including 61 with systemic lupus erythematosus, 33 with lupuslike disease, and 170 with prior adverse pregnancy outcomes. Antiphospholipid antibodies were found in 115 patients (44%), including 29 with both lupus anticoagulant and anticardiolipin antibodies (25%), one with lupus anticoagulant only (1 %), and 85 with anticardiolipin antibodies only (74%, see Samples and assays). Fifty-two of 115 patients with antiphospholipid antibodies (45%) subsequently had 59 pregnancies (group A). Of these, 19 were characterized by both lupus anticoagulant and anticardiolipin antibodies, one by lupus anticoagulant only, and 39 by anticardiolipin antibodies only. Of 149 women without anti phospholipid antibodies after screening, 54 had a subsequent pregnancy (36%, group B). In group A, 10 patients had experienced a prior thrombosis (19%) compared with four patients in group B (7%). Treatment. Treatment, including high-dose prednisone (~40 mg), was given in 11 patients. Of these, four received prednisone alone, three received prednisone plus low-dose aspirin (80 mg), and four received prednisone plus 2 X 5000 IU regular heparin subcutaneously. All but one had medium- or high-positive levels of anticardiolipin antibodies or lupus anticoagulant. Two patients with systemic lupus erythematosus in remission received prednisone for earlier activity and one as treatment of a lupus exacerbation. In all other patients treatment was given prophylactically because of the presence of anti phospholipid antibodies. Other treatments in group A included 80 mg aspirin in 16 and 2 x 5000 IU heparin given subcutaneously in three patients. In 29 pregnancies from group A no treatment was given. Treatment was given as soon as pregnancy was confirmed by ultrasonography and was continued at the same dose until 6 weeks postpartum. In patients using heparin, oral anticoagulants were given between 16 and 36 weeks of gestation instead of the heparin. In patients with medium or high levels of anticardiolipin antibodies or lupus anticoagulant allocation of treatment between 40 mg prednisone plus 80 mg aspirin versus aspirin alone, or between 40 mg prednisone plus 2 x 5000 IU subcutaneous regular heparin versus heparin alone in patients receiving oral anticoagulants because of prior thrombosis, was initially intended after randomization. However, because 28 of
Obstetric risks of anti phospholipid antibodies
27
the 32 pregnant women eligible for randomization were reluctant to use high-dose prednisone because of potential side effects, only four were randomized. Samples and assays Samples. In addition to the samples obtained for initial screening, monthly blood samples from 19 of 20 patients with lupus anticoagulant and 22 of 31 patients with medium- or high-positive levels of anticardiolipin antibodies after screening were collected during pregnancy and in the postpartum period (median eight samples, range 1 to 15). Samples from patients seen in hospitals other than in University Hospital of Utrecht were sent by mail and stored at -70 0 C within 24 hours. All assays were performed in this hospital. Blood samples for the determination of lupus anticoagulant were taken by venipuncture into a 0.1 volume of 3.8% trisodium citrate. Platelet-poor plasma was prepared by centrifuging the blood twice at 4° C at 2000g for 15 minutes. Blood samples for the preparation of serum for determinations of anticardiolipin antibodies were collected by venipuncture into glass tubes. The blood was allowed to clot at room temperature for 2 hours and centrifuged at 2000g for 10 minutes to obtain serum. Lupus anticoagulant. Three lupus anticoagulant assays were done as described previouslyl4: (1) a partial thromboplastin time with a 1 : 1 mixture of patient and pooled normal plasma with thromboplastin derived from human brain, (2) the partial thromboplastin time of test plasma on dilution of thromboplastin, and (3) kaolin clotting time index, calculated from the kaolin clotting time of control and test plasma, mixed at different ratios. Lupus anticoagulant was considered to be present when at least two assays were positive. Anticardiolipin antibodies. Anticardiolipin antibodies immunoglobulin G and immunoglobulin M were measured by enzyme-linked immunosorbent assay as described earlier 1. in a 1: 100 serum dilution. The Kingston Antiphospholipid Antibody Study Group standards were included on all enzyme-linked immunosorbent assay plates to establish cutoff levels between high-, medium-, and low-positive and normal levels, according to the second anticardiolipin standardization exercise. 1' End points. The end points of this study were fetal death, birth weight, the occurrence of small-for-gestational-age (SGA) newborns with a birth weight below the 10 percentile according to Kloosterman curves corrected for sex and parity,16 pregnancy-induced hypertension defined as a diastolic pressure of ~90 mm Hg after the twenty-fourth week of pregnancy without preexisting hypertension, and preeclampsia defined as pregnancy-induced hypertension plus proteinuria of ~0.5 gm/L. 17 Active lupus nephritis was regarded to be present when, apart from proteinuria or an elevated
28
Out et al.
July 1992 Am J Obstet Gynecol
Table I. Patient characteristics and pregnancy outcome in 59 pregnancies of patients with antiphospholipid antibodies (group A) and 54 pregnancies of patients without these antibodies (group B) Patients (No.) No. of pregnancies Systemic lupus erythematosus (n = 37) Lupuslike disease (n = 12) Prior adverse pregnancy outcome (n = 64) At least 3 spontaneous abortions At least 1 intrauterine fetal death Both spontaneous abortions and intrauterine fetal death Outcome Spontaneous abortions (n = 15) Intrauterine fetal death (n = 10) Live infant (n = 88) Gestational age* (wk, mean ± SD) Birth weight* (gm, mean ± SD) No. of children < 10th percentile* No. of pregnancies
Utrecht, The Netherlands OBJECTIVES: A prospective, controlled multicenter study was performed to estimate the obstetric risks of anti phospholipid antibodies (the lupus anticoagulant and anticardiolipin antibodies), In addition, the risks of prior thrombosis, obstetric history, systemic lupus erythematosus, and high-dose prednisone treatment were evaluated, STUDY DESIGN: After screening for anti phospholipid antibodies in patients with lupus erythematosus or women with prior fetalloss(es), 59 subsequent pregnancies with and 54 without these antibodies were followed, RESULTS: The presence of the lupus anticoagulant and a history of at least three spontaneous abortions could predict fetal loss (p = 0,032 and 0,001, respectively), In live born infants, a low birth weight could be predicted by the presence of anticardiolipin antibodies (p = 0,034), prior intrauterine fetal death (p = 0,025), and treatment with high-dose prednisone (p = 0,002). No relationships were seen between anti phospholipid antibodies and small-for-gestational-age newborns and pregnancy-induced hypertension or preeclampsia. The disappearance of anti phospholipid antibodies during pregnancy was not correlated with live birth. CONCLUSION: It is concluded that the presence of antiphospholipid antibodies is a risk factor for adverse pregnancy outcome. (AM J OBSTET GVNECOL 1992;167:26-32.)
Key words: Lupus anticoagulant, anticardiolipin antibodies, pregnancy outcome, systemic lupus erythematosus In the last decade anti phospholipid antibodies, notably the lupus anticoagulant and anticardiolipin antibodies, have received much attention. Their presence has been associated with prior fetal loss, thrombosis, and thrombocytopenia in retrospective reports. ' -3 These correlations have been described in patients with systemic lupus erythematosus, lupuslike disease, and fetal loss without underlying disease." 5 In patients with anti phospholipid antibodies, fetal loss is generally accompanied by placental thrombosis. However, the pathogenetic potential of these antibodies has not been demonstrated." There are no large-scale studies evaluating the obstetric risks for pregnant women with anti phospholipid antibodies in a prospective, controlled study. However, on the basis of studies on small series or case reports, From the Department of Obstetrics and Gynecology: the Division of Immunopathology, Department of Internal Medicine,' and the Department of Hematology,' University Hospital Utrecht. Supported by a grant from the Praeventiefonds (No. 28-1511) to HIO. Received for publication July 10, 1991,- revised December 11,1991,accepted December 30,1991. Reprint requests: HI Out, MD, Department of ObstetriC,l and Gynecology, University Hospital Utrecht, P.O. Box 85500, 3508 GA Utrecht, The Netherlands_ 611 /37391
26
it has become current practice to offer some kind of treatment. Successes have been claimed in favor of a therapy that includes high-dose prednisone," 8 but its benefits are also controversial. 9 Other successful treatments reported include subcutaneous heparin or lowdose aspirin. 9 . 10 The need for large prospective, controlled studies in pregnant women with anti phospholipid antibodies is universally recognized. II. 12 In this study we prospectively evaluated the risks of lupus anticoagulant and anticardiolipin antibodies (both level and isotype) on pregnancy outcome in 59 pregnancies complicated by anti phospholipid antibodies compared with 54 control pregnancies. We also estimated the obstetric risks of systemic lupus erythematosus, lupuslike disease, prior adverse pregnancy outcomes, history of thrombosis, and inclusion of high-dose prednisone in the treatment regImen. Material and methods
Patients. From June 1987 women who wished to conceive and who were from risk groups were screened for the presence of anti phospholipid antibodies to identify eligible patients for this study. Patients were obtained from all over the country. Risk groups included patients with systemic lupus erythematosus according
Volume 167 Number I
to American Rheumatism Association-criteria"; patients with lupuslike disease (fewer than four criteria); and patients with at least three unexplained spontaneous abortions before 12 weeks of gestation or at least one intrauterine fetal death after 12 weeks not caused by infection, abruptio placentae, apparent congenital malformations, or chromosomal abnormalities. In all patients with at least three spontaneous abortions hysterosalpingography was performed to exclude uterine malformations. A total of 264 patients were screened, including 61 with systemic lupus erythematosus, 33 with lupuslike disease, and 170 with prior adverse pregnancy outcomes. Antiphospholipid antibodies were found in 115 patients (44%), including 29 with both lupus anticoagulant and anticardiolipin antibodies (25%), one with lupus anticoagulant only (1 %), and 85 with anticardiolipin antibodies only (74%, see Samples and assays). Fifty-two of 115 patients with antiphospholipid antibodies (45%) subsequently had 59 pregnancies (group A). Of these, 19 were characterized by both lupus anticoagulant and anticardiolipin antibodies, one by lupus anticoagulant only, and 39 by anticardiolipin antibodies only. Of 149 women without anti phospholipid antibodies after screening, 54 had a subsequent pregnancy (36%, group B). In group A, 10 patients had experienced a prior thrombosis (19%) compared with four patients in group B (7%). Treatment. Treatment, including high-dose prednisone (~40 mg), was given in 11 patients. Of these, four received prednisone alone, three received prednisone plus low-dose aspirin (80 mg), and four received prednisone plus 2 X 5000 IU regular heparin subcutaneously. All but one had medium- or high-positive levels of anticardiolipin antibodies or lupus anticoagulant. Two patients with systemic lupus erythematosus in remission received prednisone for earlier activity and one as treatment of a lupus exacerbation. In all other patients treatment was given prophylactically because of the presence of anti phospholipid antibodies. Other treatments in group A included 80 mg aspirin in 16 and 2 x 5000 IU heparin given subcutaneously in three patients. In 29 pregnancies from group A no treatment was given. Treatment was given as soon as pregnancy was confirmed by ultrasonography and was continued at the same dose until 6 weeks postpartum. In patients using heparin, oral anticoagulants were given between 16 and 36 weeks of gestation instead of the heparin. In patients with medium or high levels of anticardiolipin antibodies or lupus anticoagulant allocation of treatment between 40 mg prednisone plus 80 mg aspirin versus aspirin alone, or between 40 mg prednisone plus 2 x 5000 IU subcutaneous regular heparin versus heparin alone in patients receiving oral anticoagulants because of prior thrombosis, was initially intended after randomization. However, because 28 of
Obstetric risks of anti phospholipid antibodies
27
the 32 pregnant women eligible for randomization were reluctant to use high-dose prednisone because of potential side effects, only four were randomized. Samples and assays Samples. In addition to the samples obtained for initial screening, monthly blood samples from 19 of 20 patients with lupus anticoagulant and 22 of 31 patients with medium- or high-positive levels of anticardiolipin antibodies after screening were collected during pregnancy and in the postpartum period (median eight samples, range 1 to 15). Samples from patients seen in hospitals other than in University Hospital of Utrecht were sent by mail and stored at -70 0 C within 24 hours. All assays were performed in this hospital. Blood samples for the determination of lupus anticoagulant were taken by venipuncture into a 0.1 volume of 3.8% trisodium citrate. Platelet-poor plasma was prepared by centrifuging the blood twice at 4° C at 2000g for 15 minutes. Blood samples for the preparation of serum for determinations of anticardiolipin antibodies were collected by venipuncture into glass tubes. The blood was allowed to clot at room temperature for 2 hours and centrifuged at 2000g for 10 minutes to obtain serum. Lupus anticoagulant. Three lupus anticoagulant assays were done as described previouslyl4: (1) a partial thromboplastin time with a 1 : 1 mixture of patient and pooled normal plasma with thromboplastin derived from human brain, (2) the partial thromboplastin time of test plasma on dilution of thromboplastin, and (3) kaolin clotting time index, calculated from the kaolin clotting time of control and test plasma, mixed at different ratios. Lupus anticoagulant was considered to be present when at least two assays were positive. Anticardiolipin antibodies. Anticardiolipin antibodies immunoglobulin G and immunoglobulin M were measured by enzyme-linked immunosorbent assay as described earlier 1. in a 1: 100 serum dilution. The Kingston Antiphospholipid Antibody Study Group standards were included on all enzyme-linked immunosorbent assay plates to establish cutoff levels between high-, medium-, and low-positive and normal levels, according to the second anticardiolipin standardization exercise. 1' End points. The end points of this study were fetal death, birth weight, the occurrence of small-for-gestational-age (SGA) newborns with a birth weight below the 10 percentile according to Kloosterman curves corrected for sex and parity,16 pregnancy-induced hypertension defined as a diastolic pressure of ~90 mm Hg after the twenty-fourth week of pregnancy without preexisting hypertension, and preeclampsia defined as pregnancy-induced hypertension plus proteinuria of ~0.5 gm/L. 17 Active lupus nephritis was regarded to be present when, apart from proteinuria or an elevated
28
Out et al.
July 1992 Am J Obstet Gynecol
Table I. Patient characteristics and pregnancy outcome in 59 pregnancies of patients with antiphospholipid antibodies (group A) and 54 pregnancies of patients without these antibodies (group B) Patients (No.) No. of pregnancies Systemic lupus erythematosus (n = 37) Lupuslike disease (n = 12) Prior adverse pregnancy outcome (n = 64) At least 3 spontaneous abortions At least 1 intrauterine fetal death Both spontaneous abortions and intrauterine fetal death Outcome Spontaneous abortions (n = 15) Intrauterine fetal death (n = 10) Live infant (n = 88) Gestational age* (wk, mean ± SD) Birth weight* (gm, mean ± SD) No. of children < 10th percentile* No. of pregnancies
