Correspondence
Prednisone and Aspirin in Women with Recurrent Fetal Loss
N Engl J Med 1997; 337:1629-1630November 27, 1997
- Article
To the Editor:
We applaud Laskin et al. (July 17 issue)1 for completing a large trial of prednisone and aspirin in women with autoantibodies and unexplained recurrent fetal loss. We believe, however, that several aspects of the study deserve comment.
First, 22 percent of the women had had only two pregnancy losses and therefore did not meet the traditional criteria (three or more pregnancy losses, with no more than one live birth) for the diagnosis of recurrent pregnancy loss. In addition, the women had a mixture of preembryonic, embryonic, and fetal losses. It is also unclear whether all the women were tested for other causes of pregnancy loss, although women with known conditions were excluded from the study.
Second, the proportions of women with positive tests for lupus anticoagulant were high in both the treatment and control groups (38 and 36 percent, respectively), as compared with the proportions in other studies.2 This difference raises a question about the specificity and validity of the assay used. The level of anticardiolipin-antibody positivity is not specified, and some women may have had low titers of questionable clinical importance.3 Thus, the total number of women meeting the laboratory criteria for the antiphospholipid-antibody syndrome in the study may have been very small.
Third, the autoantibodies tested were a diverse and heterogeneous group. Antinuclear antibodies have been associated with recurrent pregnancy loss in a few retrospective studies, but their presence does not increase the risk of subsequent fetal loss.4 Anti–single-stranded DNA antibodies, anti–double-stranded DNA antibodies, and antilymphocyte antibodies are not associated with pregnancy loss, and it is therefore unclear why women with these dissimilar antibodies were included.
Finally, the large number of women with repeatedly positive tests for at least one antibody (385 of the 773 women, or 50 percent) diminishes the biologic plausibility of the notion that these antibodies are of clinical importance. Is it possible that some women did not have autoimmune-mediated pregnancy loss and did not require any treatment?
The issues raised by this trial, one of the best efforts of its kind, serve to underscore the difficulty of studying autoimmune-mediated pregnancy loss and the need for uniform criteria for the diagnosis of the disorder.
4 ReferencesRobert M. Silver, M.D.
D. Ware Branch, M.D.
James R. Scott, M.D.
University of Utah Health Sciences Center, Salt Lake City, UT 841321
Laskin CA ,Bombardier C ,Hannah ME , et al. Prednisone and aspirin in women with autoantibodies and unexplained recurrent fetal loss. N Engl J Med 1997;337:148-153
Full Text | Web of Science | Medline2
Parazzini F ,Acaia B ,Faden D ,Lovotti M ,Marelli G ,Cortelazzo S . Antiphospholipid antibodies and recurrent abortion. Obstet Gynecol 1991;77:854-858
Web of Science | Medline3
Silver RM ,Porter TF ,van Leeuwen I ,Jeng G ,Scott JR ,Branch DW . Anticardiolipin antibodies: clinical consequences of “low titers.“ Obstet Gynecol 1996;87:494-500
CrossRef | Web of Science | Medline4
Harger JH ,Rabin BS ,Marchese SG . The prognostic value of antinuclear antibodies in women with recurrent pregnancy losses: a prospective controlled study. Obstet Gynecol 1989;73:419-424
Web of Science | Medline
Author/Editor Response
The authors reply:
To the Editor: Silver et al. raise important issues in interpreting the results of our study. In the group of women with two pregnancy losses, there were those who had two first-trimester losses, those who had one first-trimester and one second-trimester loss, and those who had two second-trimester losses. The obstetrical histories of these women were sufficiently abnormal to justify their inclusion in the trial.
All women were tested for other causes of pregnancy loss, including genetic causes, anatomical causes (identified on the hysterosalpingogram), and hormonal causes (identified with a timed endometrial biopsy). The women with histories of an incompetent cervix were ineligible for the study.
Lupus anticoagulant is known to be associated with recurrent pregnancy loss. The high frequency of lupus anticoagulant in our study subjects supports this finding. A positive test for anticardiolipin antibody was defined as a value at least 5 SD above the mean value in normal women. On the basis of this definition, the number of women with positive tests was relatively small.
The question of the relation between autoimmunity and fetal loss derives from observations made in women with systemic lupus erythematosus. We deliberately expanded the autoantibody screen to encompass autoantibodies present in women with this disease, because anticardiolipin antibodies and lupus anticoagulant may be only two of many markers of autoimmunity and recurrent pregnancy loss. Furthermore, we previously reported finding an increase in the prevalence of IgM antilymphocyte antibody in women with unexplained recurrent pregnancy loss.1
Silver et al. are correct in their conclusion that none of the autoantibodies, including antiphospholipid antibodies and lupus anticoagulant, directly increase the risk of fetal loss. All the studies, at best, demonstrate an association.
The comments of Silver et al. highlight the controversies inherent in the research on autoimmunity and recurrent pregnancy loss. We agree that uniform criteria for the diagnosis or classification of the disorder are needed.
1 ReferencesCarl A. Laskin, M.D.
Claire Bombardier, M.D.
Karen Spitzer, B.Sc.
University of Toronto, Toronto, ON M5S 2Z9, Canada- Citing Articles (1)
Citing Articles
1
Ronald H. W. M. Derksen, Munther A. Khamashta, D. Ware Branch. (2004) Management of the obstetric antiphospholipid syndrome. Arthritis & Rheumatism 50:4, 1028-1039
CrossRef
