1 Fertility Clinic 4071, Rigshospitalet, Copenhagen University Hospital and 2 Department of Obstetrics and Gynecology, Naestved Hospital, Naestved, Denmark
3 To whom correspondence should be addressed at: Fertility Clinic 4071, Rigshospitalet, The University Hospital of Copenhagen, Blegdamsvej 9, DK-2100 Copenhagen, Denmark. Email: obc{at}pregnancyloss.dk
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: anticardiolipin antibodies/immunotherapy in placebo-controlled trials/logistic regression analysis/prognostic variables/recurrent miscarriages
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
A recent study has shown that IgG APA directed against epitopes comprising amino acids Gly40 and Arg43 in domain I of the 2GPI have in vitro LAC activity and associate clinically strongly with thrombosis (de Laat et al., 2004
). Most of or all IgM APA and the subset of IgG APA without specificity against this epitope seem to be without LAC activity and clinical association to thrombosis.
Several attempts have been made to quantify the impact of APA on pregnancy outcome but only a limited number of these studies have been prospective. The patients included in the prospective studies have been recruited from very different populations: women with systemic lupus erythematosus (Ginsberg et al., 1992; Out et al., 1992
), healthy women with low risk pregnancies (Lockwood et al., 1989
; Infante-Rivard et al., 1991
; Pattison et al., 1993
; Lynch et al., 1994
; Yasuda et al., 1995
), women with recurrent miscarriage (RM) (Barbui et al., 1988
; Parazzini et al., 1991
; Rai et al., 1995b
) or mixtures of these populations. In the prospective studies, the effect of ACA on live birth rate has been reported very differently with some studies finding no effect (Taylor et al., 1990
; Infante-Rivard et al., 1991
; Out et al., 1992
; Melk et al., 1995
, Simpson et al., 1998
), and other studies reporting a great negative impact of these antibodies (Tulppala et al., 1993
; Yasuda et al., 1995
; Rai et al., 1995b
).
The different results in the prospective studies could partly be attributed to poorly standardized ACA tests (Harris et al., 1987b) reflected in a significant interlaboratory variation in the detection of ACA (Peaceman et al., 1992
; Roberts et al., 2002
), significant fluctuations of the antibody levels over time in the same patient (Rai et al. 1995a
; Donohoe et al. 2002
) and very different cut off levels (Quenby and Farquharson, 1993
; Tulppala et al., 1993
; Rai et al., 1995 a
,b
). Another important reason for the different estimates of risk conferred by ACA may be that the patient populations monitored have been very heterogeneous with regard to underlying disease, the number of miscarriages, age and other important prognostic factors. Few authors have adjusted for the influence of other prognostic variables by using logistic regression analysis (Lockwood et al., 1989
; Infante-Rivard et al., 1991
; Out et al. 1992
, Lynch et al., 1994
). In neither of the studies did the included patients suffer from RM. Finally, very few of the published prospective studies have focused especially on RM patients with exclusively ACA (Triplett, 1989
), in most prospective studies ACA and LAC positive patients have been considered together.
For a clinician who is treating patients with RM, previous studies do not adequately answer the question: what is the prognostic value of an elevated ACA titre in the absence of LAC in non-anticoagulation treated RM patients with regard to live birth rate and the development of late pregnancy complications such as preeclampsia, low birth weight and preterm labour? This is the question we try to answer in this study.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
In the ALT trial, exclusion criteria were positivity for lymphocytotoxic antibodies, LAC, or high titres of antinuclear antibodies (ANA) or anti-double-stranded deoxyribonucleic acid (-dsDNA) antibodies.
The IvIg trials comprised two trials where, in the first trial, patients with at least three previous miscarriages and mainly secondary RM were included whereas, in the second, only patients with a minimum of four previous miscarriages with both primary and secondary RM were included. For both IvIg trials no immunological or coagulation-related exclusions were made (Christiansen et al. 2002). Two patients included in the trials had a positive LAC test and were not included in the present analysis.
Eighty-four patients (57%) in the total study population had had primary RM, whereas 63 (43%) had suffered secondary RM having carried at least one pregnancy (live birth/stillbirth) beyond the 28th week of gestation before the sequence of miscarriages. Only 26 (4.1%) of the patients 625 pregnancy losses had occurred after gestational week 10 (and before week 22). Table I displays background parameters for the patients. There were statistically significant differences between the ALT trial and IvIg trial with regard to the number of previous miscarriages, the frequency of secondary RM and the frequency of positivity for ACA, partly due to the different inclusion and exclusion criteria in these trials. However, no significant differences in the frequencies of these parameters were found between the treatment and placebo groups, illustrating that the allocation to the two groups had indeed been random.
|
IvIg trial
In the first study IvIg or placebo (albumin) 0.4 g/kg bodyweight was infused from 5th week to 8th week and afterwards every second week until week 34. In the second trial 0.8 g/kg bodyweight was infused weekly from 5th week to 10th week and afterwards every fortnight until week 26.
In neither of the trials did any of the patients receive prednisone or anticoagulation therapy such as heparin or aspirin. No hormonal support (progesterone, hCG) was given during pregnancy and all patients conceived spontaneously.
Laboratory analyses
Screening for LAC was performed by measuring the activated partial thrombin time (APTT) in plasma according to the international consensus from the period the study was done. It recommended APTT as a test that alone or in conjunction with other coagulation tests should be the first step in investigation for LAC (Wilson et al., 1999). Thrombocyte-poor citrated plasma from freshly drawn blood was used for the APTT determination. The APTT measurement was undertaken with the Automated APTT kit (Organon Teknika, NC). One hundred microlitres of plasma was incubated with 100 µl of rabbit brain phospholipids together with micronized silica as activator and a buffer for 5 min at 37 °C. One hundred microlitres of 20 mM CaCl2 was added and the clotting time was measured optically with a Behring Coagulation Timer. As a control, the clotting time of samples from a pool of normal plasma was included in every analysis. If APTT was more than 40 s, a confirmatory test for LAC was performed using commercially available kits (Viper Quick LA-test and Viper Quick LA-check; Organon Technics, Durham, NC).
Details concerning the laboratory tests for ANA and -dsDNA have been described previously (Christiansen et al., 1992
, 2002
).
Detection of anticardiolipin antibodies
The first blood sample (sample I) was taken before pregnancy in the ALT trial. In the IvIg trials, sample I was taken as soon as a pregnancy test was positive at the latest 5 days after the first day of the missed period. In both the ALT and IvIg trial, sample II was taken between the 6th and l0th gestational week calculated from the last menstrual period in the patient's index pregnancy; in the ALT trial this was the first subsequent pregnancy after sample I was taken. Among the ACA positive patients in this study, one patient had a 2 week interval between the samples, five had 4 weeks, and three had 5 weeks, whereas the remaining patients had 6 weeks between the two samples.
Samples from patients and controls were allowed to coagulate. Serum was separated by centrifugation, stored at 80 °C, and was kept frozen before use. The ACA analyses were made after the trials were concluded using an ELISA assay. The kit (Varelisa; Elias, Freiburg, Germany) uses the principle of a transferable solid phase. The wells of the ELISA plate are coated with bovine phospholipids complexed with 2-GPI using a non-ethanol based water soluble coating buffer in order to preserve the 3-dimensional structure of the cardiolipin molecule. The tested samples were diluted 1:100 with a phosphate buffer supplemented with 10% bovine serum albumin and 1% sodium azide. The kit fulfils the recommendations of Harris et al. (1987a)
and the assays were carried out against an established standard serum from the Rayne Institute, London. Photometric quantification at 492 nm and correction for background activity at 620 nm were undertaken, and for the calculation of unknown samples, mean optical densities were compared with the computerized standard curve. The results for immunoglobulin (Ig)M and (Ig)G class ACA were calibrated in M-phospholipid (MPL)-U/ml and G-phospholipid (GPL)-U/ml, respectively. For each test kit, one control serum and seven standard samples in concentrations ranging from 0 to 60 MPL-U/ml and from 0 to 100 GPL-U/ml were included. All samples were determined in duplicate, and samples I and II from the same patients were always assayed on the same ELISA plate to avoid the effect of any inter-assay variation. The intra-assay coefficient of variation in this study was in our laboratory found to be <1% for low values (IgM <2 MPL-U/ml, IgG<4 GPL-U/ml), 2.5% for mid-range values (IgM 27 MPL-U/ml, IgG 410 GPL-U/ml) and 4% for high values (IgM
7 MPL-U/ml and IgG
10 GPL-U/ml). The inter-assay coefficients of variation for this kit have been reported to be 7.810.4% for IgG ACA and 18.6% for IgM ACA (Wong et al., 2004
). Coefficients of variation below 20% are considered appropriate (Wong et al., 2004
).
Establishment of cut-off levels
The impact of a repeatedly positive ACA value on future pregnancy outcome was estimated using two different sets of cut-off levels. The first set was defined as the upper 95th percentiles of ACA measurements carried out on the same ELISA plates as used in this study in a population of 50 Danish non-pregnant women of fertile age (Christiansen et al., 1992). The normal population-based cut-offs were 7.0 MPL-U/ml and 22.0 GPL-U/ml for IgM and IgG ACA, respectively. If a patient was repeatedly positive for IgM ACA, IgG ACA or both of them according to these cut-off levels she was designated IgM ACA-NP+ (normal population positive), IgG ACA-NP+ or ACA-NP+.
The second set of cut-off values was established as the 10% percentiles (decentiles) above which the patients had a significant decline in live birth rate in the present prospective study. Patients being repeatedly positive according to these cut-off levels would be designated IgM ACA-RM+ (recurrent miscarriage population positive), IgG ACA-RM+ or ACA-RM+.
Statistics
Differences of background variables between the treatment and the placebo groups, between the ALT and the IvIg trials and differences between subsequent birth rates and the presence/absence of ACA, were tested with the 2 test.
Heterogenicity of live birth rates between the six different ACA-RM+ subgroups in the study (ACA-RM+ patients in the ALT trial and the first and second IvIg trial allocated to treatment or placebo) was tested by the 2 test. The same heterogeneity test was done for the patients in the six ACA-RM groups.
The outcome measure of live birth versus miscarriage (odds for live birth) was studied by a stepwise, backward, multivariate logistic regression analysis performed to identify the possible prognostic variables that best predicted the outcome of live birth. Data analyses were performed using the SPSS (Statistical Package for Social Sciences) version 10.0. The following variables were entered into the model: the number of previous miscarriages, the age of the patient, treatment status, the presence or absence of other autoantibodies (ANA, -dsDNA) and the ACA test status. Age and number of previous miscarriages were entered as continuous variables whereas the remaining variables were divided into binary categories. The Odds Ratios (ORs) in the logistic regression analysis were calculated as the ratios between the odds for live birth in patients positive for an independent variable and the odds for live birth in patients negative for the variable with or without adjustment of the impact of the other variables. With regard to the continuous variables, the ORs refer to the ratios between the odds for live birth in the group with a higher number of previous miscarriages or age and the group with the number of miscarriages or age just below. Using backward elimination we excluded variables that were not significantly associated with the outcome measure. The relevant two-factor interaction term defined as the product of treatment and ACA positivity was tested in the model. The OR for the interaction variable is the odds for live birth in treated ACA positive patients divided by the odds for live birth in all other patients (untreated ACA positive patients and treated and untreated ACA negative patients).
Testing the squared values of the continuous variables age and number of miscarriages confirmed the adequacy of the model.
Fisher's exact test and independent samples test were used to investigate for differences between ACA positive and ACA negative patients with regard to late pregnancy complications in those patients who gave birth to a child.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Cut-off values predicted by this study
Figure 1 shows the live birth rates in this study when the patients are divided into decentile groups according to the mean (sample I and II) of IgM ACA and IgG ACA values. For IgM ACA values higher than 4.6 MPL-U/ml (separating the 6th and 7th decentile groups) and IgG values higher than 12.8 GPL-U/ml (separating the 8th and 9th decentile groups), the live birth rates were in general lower than in lower decentile groups and below 52% and these values were chosen as the cut-off levels for ACA (ACA-RM values) derived from this prospective study. In order to be designated ACA-RM+, a patient thus had to have IgM ACA-RM4.6 MPL-U/ml and/or IgG ACA-RM
12.8 GPL-U/ml in both samples I and II.
|
ACA tests
Among all 147 study patients, 60 (41%) were ACA-RM+, 47 (32%) were IgM ACA-RM+, 19 (13%) were IgG ACA-RM+ and 31(21%) were ACA-NP+. With regard to the samples taken in the ALT trial we have previously reported that ALT therapy does not change ACA concentrations (Christiansen et al., 1992). With regard to the samples in the IvIg trials, the median concentration of IgM ACA in sample I was 4.2 MPL-U/ml and 4.4 MPL-U/ml in the placebo and treatment groups, respectively, and the corresponding concentrations were 4.9 and 5.0 MPL-U/ml for sample II. The median concentration of IgG ACA in sample I was 6.2 GPL-U/ml and 6.0 GPL-U/ml in the placebo and IvIg treatment groups, respectively, and the corresponding concentrations were 6.0 and 8.5 GPL-U/ml in sample II. Neither of the median concentrations was statistically significantly different between the placebo and treatment groups.
Table I shows the distribution of ACA-RM+ and ACA-NP+ patients according to inclusion in either of the two trials and allocation groups. The frequencies of a positive test were not statistically significantly different between patients in the treatment and placebo groups.
In Tables II and III are shown the live birth rates according to ACA status stratified for number of previous miscarriages and age, respectively. There was a decrease in live birth rate with increased number of previous miscarriages and increased age for both ACA-RM+ and ACA-NP+. The live birth rate tended at each stratum of numbers of miscarriages and age to be lower in the ACA positive groups. There was a statistically significant correlation between the number of previous miscarriages and the frequency of ACA positivity in the patients regardless of which cut off value was used (Table II), whereas the correlation between ACA positivity and increasing age was not significant (Table III).
|
|
|
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The present study showed that the OR for live birth in repeatedly ACA positive, LAC negative RM patients was 0.36 (0.20.7), P=0.006 using cut-off values established in the study (ACA-RM values) and adjusting for the impact of confounding variables (Table IV). The results point at IgM ACA as a more sensitive negative predictor of live birth than IgG ACA. Using cut-off values derived from a normal female population of fertile age (ACA-NP values), the OR for live birth in ACA positive patients was borderline significant before adjustment for prognostic variables but became non-significant after this adjustment resulting in an OR of 0.48 (0.21.1), P=0.1.
The 60 repeatedly ACA-RM+ patients in the study comprise, to our knowledge, the greatest number of non-anticoagulation-treated ACA positive, LAC negative patients with RM followed prospectively into the next pregnancy being twelve times the number of similar patients followed by Rai et al. (1995b). We believe that the number of ACA positive and negative patients is sufficient for making a thorough analysis, adjusting for the possible confounders by logistic regression, and the estimate of the prospective impact of the antibodies is thus the most exact so far published.
Since heparin and aspirin therapy has gained widespread useand in many centres is considered mandatory for ACA positive patientsno results of pregnancy outcome in new cohorts of non-anticoagulation treated ACA positive RM women is expected in the future. The widespread use of heparin/aspirin therapy is due to the results from two (non-blinded) trials that found a statistically significantly higher live birth rate in APA positive RM patients receiving both heparin and low dose aspirin compared with low dose aspirin alone (Kutteh, 1996; Rai et al., 1997
). In the latter trial almost exclusively patients with LAC were included.
In our study, the chance for live birth in ACA positive RM patients (3940%) is much higher than the 10% live birth rate very often referred to as reflecting the live birth rate in untreated RM patients with APA (Rai et al., 1995b). This prospective study compared 20 RM patients who were positive for APA and none of whom received any therapyamong these only two (=10%) gave birth to live children. However, in this study as many as 15 of the patients (75%) were positive for LAC and only five patients were exclusively ACA positive, indicating that the cohort is not typical for the general RM population.
As pointed out in casecontrol studies (Lockshin et al., 1989; Gris et al., 2000
) and laboratory studies (de Laat et al., 2004
) there is a difference between patients identified by the presence of LAC and patients identified by the presence of ACA. The prevalence of ACA in the RM population is generally reported to be much higher than the prevalence of LAC (Cowshock et al., 1986
; Unander et al., 1987
; Lockwood et al., 1989
) and the concordance between LAC and ACA has been reported to vary from 36 to 68% for RM patients (Rote et al., 1990
). This was confirmed in the present study since in the original patient population before exclusions were undertaken, only five were LAC positive whereas 65 were ACA-RM+ (including the five LAC positive patients). Among the 60 patients who were ACA-RM+ and LAC negative, none had experienced any thrombotic events whereas three out of the five LAC positive patients had a history of venous thrombosis, indicating that our screening for LAC was indeed quite specific with regard to identifying patients with an increased risk of thrombosis.
Only once before has discrimination between LAC and ACA been done in prospective studies of pregnancy outcome. Out et al. (1992) discriminated between ACA and antibodies with LAC activity in their prospective study of patients with autoimmune disease and repeated pregnancy losses, finding that only the presence of LAC and not ACA was predictive of a miscarriage in the next pregnancy. Interestingly, when discrimination is done between LAC and ACA in the prospective study by Rai et al. (1995b)
the live birth rate was 0% in RM patients with LAC and 40% in those with ACA alonealso providing evidence that the presence of LAC has a much greater negative impact than ACA. We therefore find it important, as done in the present study, to evaluate the influence of ACA alone in the absence of LAC.
In prospective studies, adjustment for the influence of confounders, which can affect the main outcome variable, should always be undertaken when the impact of an independent variable is estimated. The importance of adjusting for other prognostic variables when evaluating the real impact of ACA on live birth rate is evidenced in Tables II and III, showing that the frequency of ACA positive patients increases with the number of previous miscarriagesa well-established negative prognostic factor for live birth rate (Cauchi et al., 1995). The number of miscarriages variable is therefore meeting the criteria of being a confounder in this study, since it is significantly associated to both ACA and live birth rate. It cannot be ruled out that the production of ACA may be the result of the previous carriages of a dead fetus in the uterus.
A negative impact on pregnancy outcome of non-APA autoantibodies has often been reported (Gleicher and el-Roeiy, 1988; Tulppala et al., 1993
; Matzner et al., 1994
). Since non-APA autoantibodies are often found associated with ACA we also found it important to control for the presence of these antibodies.
Our logistic regression analysis indicated that when adjustment for the effect of the variables selected to be of importance (Table IV) is carried out, the impact of presence of ACA defined by a normal population-derived cut-off value on pregnancy outcome becomes statistically insignificant. However, no influence of the confounder control on the impact of the patient-derived cut-off value could be seen.
Although few patients in our study were positive for ANA and -dsDNA, these autoantibodies remained as significant negative predictors of live birth after adjustment. This is in line with Gleicher's hypothesis of the reproductive autoimmune failure syndrome (Gleicher and el-Roeiy, 1988
): that a series of autoantibodies can be found with increased prevalence in RM patients as a sign of a more wide-spread autoimmune response in these patients. It is possible that the presence of ACA in the absence of LAC, together with the presence of other autoantibodes, is a marker for a general tendency to autoreactivity, which is prognostically bad for pregnancy although the autoantibodies, including ACA, might per se be epiphenomena.
Our study is the first prospective study with confounder control in a population of exclusively RM patients. Out et al. (1992) adjusted for confounders in their study in a mixed population of patients with lupus erythematosus or with one or more prior fetal losses and found that a new miscarriage was predicted by the presence of LAC or a history of RM but not the ACA level. Lynch et al. (1994)
in a population of healthy pregnant women found a significant association of IgG ACA and pregnancy loss after adjustment for confounders.
There are a number of potential limitations to our study. In the study were included patients who participated in three previously published randomized placebo-controlled trials of immunotherapy with allogeneic lymphocytes or IvIg and it cannot totally be ruled out that some of the treatments influence pregnancy outcome in ACA positive patients. An analysis limited to the placebo arms of the studies might have been preferable but since it would then be restricted to much fewer patients the estimate of the OR for livebirth in ACA-RM+ patients would have wide 95% confidence limits: 0.22.0 (Table V), which is too imprecise to be of clinical value. We believe that patients in the treated groups can be included in the study due to the following arguments: (i) assignment to the various treatments was independent of whether the patients were positive of ACA or not, (ii) neither of the treatments affects the ACA titres (Christiansen et al., 1992; this study), (iii) there was no evidence that the treatments have a selective effect in ACA positive patients since the two-factor interaction term between treatment and ACA-positivity was not statistically significant in the backward elimination procedure in the logistic regression analysis, (iv) tests for heterogeneity did not find the lymphocyte-immunized, IvIg-treated and the placebo groups significantly different with regard to live birth rates in ACA positive and negative patients. Patient groups in different trials can be combined in meta-analyses if no statisticallly significant heterogeneity of the trials with regard to the main outcome measure can be demonstrated and we therefore find it justified combining patients receiving different treatments. In addition, any other biases in age, the numbers of previous miscarriages etc. between the different groups were adjusted for in the logistic regression analysis.
The negative impact on pregnancy outcome was in our study primarily confined to the presence of IgM ACA and not IgG ACA, although we admit that there were only few patients in the latter group.
In contrast to this, some studies have pointed towards IgG ACA as having the highest correlation to pregnancy loss (Lynch et al., 1994). The hypothesis behind the claim that IgM ACA is a less sensible predictor of miscarriage is that a high incidence of IgM ACA has been found in the setting of infectious diseases (Unander et al., 1987
), however, none of our patients had had infectious disease at the time of the blood samples or in the 3 month period before. The role of the different isotypes of ACA in RM must be elucidated in further prospective studies.
Our study is the first prospective study in RM patients where all patients designated to be ACA+ had repeatedly positive tests. It is recommended that due to the fluctuation of ACA, emphasis should only be put on an ACA test that remains positive in two different blood samples taken with at least 6 weeks interval (Wilson et al., 1999). The usefulness of this criterion for making ACA studies more homogeneous can be questioned as long as there is no common definition of the cut-off values and no criteria for the tests' dating in relation to pregnancy. Since ACA are fluctuating, a positive test taken during or immediately before pregnancy must logically be of greater prognostic value than a test taken months before. In the present study, we made efforts to ensure that the ACA tests really were relevant for the next pregnancy since all patients designed as ACA positive were positive during early pregnancy and cut-off values obtained by two different methods were used.
The Varelisa test kit used for detection of ACA fulfilled the requirements for detection of clinically important ACA (Galli et al., 2003): the wells contained
2-GPI, the buffer contained bovine serum albumin and the intra-assay and inter-assay variability coefficients were appropriate (Wong et al., 2004
).
The cut-off levels used in the present study based on the actual outcome (ACA-RM values) need confirmation in future studies since they were (post-hoc) selected to be the cut-off values with the highest specificity for pregnancy loss in this study but this may not hold true in another study. Indeed if the ACA-RM cut-off levels are applied on our control population of healthy females of fertile age, 30% would be positive for IgM and/or IgG ACA so it is not obvious which of the two cut-off levels tested in this study is superior. The ACA-NP cut-off values may be the best for differentiating between women who are at risk for developing RM and those who are not, whereas the ACA-RM cut-offs may be best for predicting pregnancy outcome in patients who have already got the RM diagnosis.
Several studies have reported that the risk of fetal growth retardation and pre-eclampsia is increased in ACA positive women (Branch, 1992; Yasuda et al., 1995
; Kutteh et al., 1996
). In the present study no differences were found concerning pregnancy complications and ACA status (Table VI), however, it must be emphasized that the study was made primarily to assess the impact of ACA on live birth rate and had not enough statistical power to evaluate the frequency of late pregnancy complications.
In conclusion, our study and other prospective studies (Taylor et al., 1990; Melk et al., 1995
) suggest that the real impact of ACA on pregnancy outcome in RM in the absence of LAC is less than generally believed. In the present study the most likely chance for live birth in ACA positive patients is 3648% of the chance of ACA negative patients. However, the 95% confidence intervals indicate that the real chance of live birth could range from as low as 20% to as high as 70% of that of ACA negative patients. Factors responsible for our findings are probably the focus on patients negative for LAC and the adjustment for the influence of other prognostic factors.
We believe that the results stress the urgent need for real placebo-controlled trials of anticoagulation therapy and other forms of therapy in LAC negative, ACA positive patients, taking into account that one of the two heparin/aspirin trials included almost exclusively the minority of patients that were LAC positive (Rai et al., 1997). If these trials are not carried out, dozens of women will daily be exposed to a treatment with potential risks (although low) of osteoporosis, and/or haemorhage (Ruiz-Irastorza et al., 2001
), together with economic costs, which might have less therapeutic effect than previously believed because the spontaneous prognosis in ACA positive, LAC negative RM patients seems to be fair.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Branch DW (1992) Outcome of treated pregnancies in women with APA syndrome: an update of the Utah experience. Obstet Gynecol 80, 614620.[Abstract]
Cauchi MN, Coulam CB, Cowchock S et al. (1995) Predictive factors in recurrent spontaneous abortersa multicenter study. Am J Reprod Immunol 33, 165170.[ISI][Medline]
Christiansen OB, Lauritsen JG, Andersen ES and Grunnet N (1989) Autoimmune associated recurrent abortions. Hum Reprod 4, 913917.[Abstract]
Christiansen OB, Christiansen BS, Husth M, Mathiesen O, Lauritsen JG and Grunnet N (1992) Prospective study of anticardiolipin antibodies in immunized and untreated women with recurrent spontaneous abortions. Fertil Steril 58, 328334.[ISI][Medline]
Christiansen OB, Mathiesen O, Husth M, Rasmussen KL, Ingerslev HJ, Lauritsen JG and Grunnet N (1995) Placebo-controlled trial of treatment of unexplained secondary recurrent spontaneous abortions and recurrent late spontaneous abortions with i.v. Immunoglobulin. Hum Reprod 10, 26902695.[Abstract]
Christiansen OB, Pedersen B, Rosgaard A and Husth M (2002) A randomized, double blind, placebo-controlled trial of intravenous immunoglobulin in the prevention of recurrent miscarriage: Evidence for a therapeutic effect in women with secondary recurrent miscarriage. Hum Reprod 3, 809816.[CrossRef]
Cowshock S, Smith J and Gocial B (1986) Antibodies to phospholipids and nuclear antigens in patients with repeated abortions. Am J Obstet Gynecol 155, 10021010.[ISI][Medline]
de Laat B, Derksen RHWM, Urbanus RT and de Groot PG (2004) IgG antibodies that recognize epitope Gly40-Arg43 in domain I of b2-glycoprotein I cause LAC and their presence correlates strongly with thrombosis. Blood prepublished online October 26, DOI 10.1182/blood-2003-09-3387.
Donohoe S, Quenby S, Mackie I, Panal G, Farquharson R, Malia R, Kingdom J and Machin S (2002) Fluctuations in levels of antiphospholipid antibodies and increased coagulation activation markers in normal and heparin-treated antiphospholipid syndrome pregnancies. Lupus 11, 1120.[CrossRef][ISI][Medline]
Galli M and Barbui T (2003) Antiphospholipid antibodies and pregnancy. Best Pract Res Clin Haematol 16, 211225.[CrossRef][Medline]
Galli M, Luciani D, Bertolini G and Barbui T (2003) Anti-2-glycoprotein I, antiprothrombin antiboidies, and the risk of thrombosis in the antiphospholipid syndrome. Blood 102, 27172723.
Ginsberg JS, Brill-Edwards P, Johnston M, Denburg JA, Andrew M, Burrows RF, Bensen W, Cividino A and Long AA (1992) Relationship of antiphospholipid antibodies to pregnancy loss in patients with systemic lupus erythematosus: a cross-sectional study. Blood 80, 975980.[Abstract]
Gleicher N and el-Roeiy A (1988) The reproductive autoimmune failure syndrome. Am J Obstet Gynecol 159, 223227.[ISI][Medline]
Gris JC, Quere I, Sanmarco M et al. (2000) Antiphospholipid and antiprotein syndromes in nonthrombotic, non-autoimmune women with unexplained recurrent primary early foetal loss. The Nimes Obstetricians and Haematologists Study NOHA. Thromb Haemost 84, 228236.[ISI][Medline]
Harris EN, Gharavi AE, Patel SP and Hughes GRV (1987a) Evaluation of the anti-cardiolipin test: report of an international workshop held 4th of April 1986. J Exp Immunol 68, 215222.
Harris EN, Gharavi AE and Hughes GRV (1987b) Anti-cardiolipin antibody testing: The need for standardization. Arthritis Rheum 30, 835837.[ISI][Medline]
Harris EN (1995) The antiphospholipid syndrome. Diagnosis, management and pathogenesis. Clin Rev Allergy Immunol 13, 3948.[ISI][Medline]
Infante-Rivard C, David M, Gauthier R and Rivard GE (1991) Lupus anticoagulants, anticardiolipin antibodies, and fetal loss. A case-control study. N Engl J Med 325, 10631066.[Abstract]
Kutteh WH (1996) Antiphospholipid antibody-associated recurrent pregnancy loss: treatment with heparin and low-dose aspirin is superior to low-dose aspirin alone. Am J Obstet Gynecol 174, 15841589.[ISI][Medline]
Laskin CA, Bombardier C, Hannah ME et al. (1997) Predisnone and aspirin in women with autoantibodies and unexplained fetal loss. N Eng J Med 337, 148153.
Lockwood CJ, Romero R, Feinberg RF, Clyne LP, Coster B and Hobbins JC (1989) The prevalence and biological significance of lupus anticoagulant and anticardiolipin antibodies in a general obstetric population. Am J Obstet Gynecol 160, 439443.[ISI][Medline]
Lockshin MD, Druzin ML and Qamar T (1989) Prednisone does not prevent recurrent fetal death in women with antiphospholipid antibody. Am J Obstet Gynecol 160, 439443.[ISI][Medline]
Lynch A, Marlar R, Murphy J, Davila G, Santos M, Rutledge J and Emlen W (1994) Antiphospholipid antibodies in predicting adverse pregnancy outcome. A prospective study. Ann Intern Med 120, 470475.
Maier DB and Parke A (1989) Subclinical autoimmunity in recurrent aborters. Fertil Steril 51, 280285.[ISI][Medline]
Matzner W, Chong P, Xu G and Ching W (1994) Characterization of anti phospholipid antibodies in women with recurrent spontaneous abortions. J. Reprod Med 39, 2730.[ISI][Medline]
Melk A, Mueller-Eckhardt G, Polten B, Lattermann A, Heine O and Hoffmann O (1995) Diagnostic and prognostic significance of anticardiolipin antibodies with recurrent spontaneous abortions. Am J Reprod Immunol 33, 228233.[ISI][Medline]
Out HJ, Bruinse HW, Christianens GCML, Vliet MV, Groot PG, Nieuwenhuis HK and Derksen RHWM (1992) A prospective controlled multicenter study on the obstetric risks of pregnant women with antiphospholipid antibodies. Am J Obstet Gynecol 167, 2632.[ISI][Medline]
Parazzini F, Acaia B, Faden D, Lovotti M, Marelli G and Cortelazzo S (1991) Antiphospholipid antibodies and recurrent abortion. Obstet Gynecol 77, 854858.[Abstract]
Parke AL, Wilson D and Maier D (1991) The prevalence of anti phospholipid antibodies in women with recurrent spontaneous abortion, women with successful pregnancies, and women who have never been pregnant. Arthritis Rheum 34, 12311235.[ISI][Medline]
Pattison NS, Chamley LW, McKay EJ, Liggins GC and Butler WS (1993) Antiphospolipid antibodies in pregnancy: prevalence and clinical associations. Br J Obstet Gynaecol 100, 909913.[ISI][Medline]
Peaceman AM, Silver RK, MacGregor SN and Socol ML (1992) Interlaboratory variation in antiphospholipid antibody testing. Am J Obstet Gynecol 166, 17801787.[ISI][Medline]
Petri M, Golbus M, Andersson R, Whiting-O'Keefe Q, Corash L and Hellmann D (1987) Antinuclear antibody, lupus anticoagulant, and cardiolipin antibodies in women with idiopathic habitual abortion. Arthritis Rheum 30, 601606.[ISI][Medline]
Quenby SM and Farquharson RG (1993) Predicting recurring miscarriage: what is important? Obstet Gynecol 82, 132138.[Abstract]
Rai RS, Regan L, Clifford K, Pickering W, Dave M, Mackie L, McNally T and Cohen H (1995a) Antiphospholipid antibodies and 2-g1ycoprotein I in 500 women with recurrent miscarriage: Results of a comprehensive screening approach. Hum Reprod 10, 20012005.[Abstract]
Rai RS, Clifford K, Cohen H and Regan L (1995b) High prospective fetal loss rate in untreated pregnancies of women with recurrent miscarriage and antiphospholipid antibodies. Hum Reprod 10, 33013304.[Abstract]
Rai R, Cohen H, Dave M and Regan L (1997) Randomised controlled trial of aspirin and aspirin plus heparin in pregnant women with recurrent miscarriage associated with phospholipid antibodies (or antiphospholipid antibodies). BMJ 314, 253257.
Roberts JM, Macara LM, Chalmers EA and Smith GCS (2002) Inter-assay variation in antiphospholipid antibody testing. Br J Obstet Gynaecol 109, 348349.
Rote NS, Dostal-Johnson D and Branch DW (1990) Antiphospholipid antibodies and recurrent pregnancy loss: correlation between the activated partial thromboplastin time and antibodies against phosphatidylserine and cardiolipin. Am J Obstet Gynecol 163, 575584.[ISI][Medline]
Ruiz-Irastorza G, Khamashta MA, Nelson-Piercy C and Hughes GR (2001) Lupus pregnancy: is heparin a risk factor for osteoporosis? Lupus 10, 597600.[CrossRef][ISI][Medline]
Simpson JL, Carson SA, Chesney C et al. (1998) Lack of association between antiphospholipid antibodies and first-trimester spontaneous abortion: prospective study of pregnancies detected within 21 days of conception. Fertil Steril 69, 814820.[CrossRef][ISI][Medline]
Stephenson MD (1996) Frequency of factors associated with habitual abortion in 197 couples. Fertil Steril 66, 2429.[ISI][Medline]
Taylor M, Cauchi MN and Buchanan RRC (1990) The lupus anticoagulant, anticardiolipin antibodies and recurrent miscarriage. Am J Reprod Immunol 23, 3336.[ISI][Medline]
Triplett DA, Brandt JT, Musgrave KA and Orr CA (1988) The relationship between lupus anticoagulants and antibodies to phospholipid. JAMA 259, 550554.[Abstract]
Triplett DA (1989) Antiphospholipid Antibodies and Recurrent Pregnancy Loss. Am J Reprod Immunol 20, 5267.[ISI][Medline]
Tulppala M, Palo ST, Ramsay T, Miettinen A, Salonen R and Ylikorkala O (1993) A prospective study of 63 couples with a history of recurrent spontaneous abortion: contributing factors and outcome of subsequent pregnancies. Hum Reprod 8, 764770.[Abstract]
Unander AM, Norberg R, Hahn L and Årfors L (1987) Anticardiolipin antibodies and complement in ninety-nine women with habitual abortion. Am J Obstet Gynecol 156, 114119.[ISI][Medline]
Wilson WA, Gharavi AE, Koike T et al. (1999) International consensus statement on preliminary classification criteria for definite antiphospholipid syndrome: report of an international workshop. Arthritis Rheum 42, 309311.[CrossRef]
Wong RCW, Favaloro EJ, Pollock W et al. (2004) A multi-centre evaluation of the intra-assay and inter-assay variation of commercial and in-house anti-cardiolipin antibody asays. Pathology 36, 182192.[CrossRef][ISI][Medline]
Yasuda M, Takakuwa K, Tokunaga A and Tanaka K (1995) Prospective study of the association between anticardiolipin antibody and outcome of pregnancy. Obstet Gynecol 86, 555559.
Submitted on May 4, 2004; resubmitted on December 21, 2004; accepted on January 12, 2005.
|