Background
The link between blood hypercoagulability and infertility or in vitro fertilization (IVF) failure is a puzzling issue. Hypercoagulability could be intrinsic or caused by the hormone treatment preceding the IVF procedure [
1‐
6]. Tissue factor is the major trigger of blood coagulation and thrombin generation is the ultimate step that leads to fibrin formation. The activation of coagulation induced by tissue factor (TF) expressed by perivascular decidualized human endometrial stromal cells is an essential part of the mechanism that favors blastocyst implantation and prevents peri-implantational hemorrhage during endovascular trophoblast invasion. Thrombin generation is required for cell proliferation, neoangiogenesis, trophoblast invasion and remodeling of the spiral arteries and arterioles [
7‐
9]. Thus, the shift of blood coagulation equilibrium towards locally enhancement of thrombin generation may have some beneficial effects for a positive outcome of IVF. On the other hand, in infertile women activation or dysfunction of platelets, endothelial cells and monocytes has been observed [
10,
11].
Newer laboratory assays allow the assessment of global blood coagulation and clot formation process. Among them, the Calibrated Automated Thrombogram® and the minimal TF-triggered whole blood thromboelastometry allow the evaluation of thrombin generation and clot formation processes [
12‐
14]. The measurement of the procoagulant phospholipid dependent clotting time (Procoag-PPL®) reflects the plasma concentration of procoagulant membrane vesicles of cellular origin [
15,
16]. Biomarkers of endothelial cell activation such as thrombomodulin activity (TMa) and TF activity (TFa) measured in plasma offer information on the status of the endothelial cells at the vasculature.
The aim of the present prospective, observational longitudinal study was to identify biomarkers of hypercoagulability which could have some predictive value for the IVF outcome. Thrombin generation, clot formation kinetics and molecular biomarkers of cellular hypercoagulability were assessed at women eligible for IVF at baseline (before any hormone treatment administration), at the down-regulation phase of the menstrual cycle and after ovarian stimulation. Biochemical diagnosis of pregnancy was the end-point of the study.
Methods
Study design and participants
A monocentric prospective, non-interventional cohort study was designed. From June 2014 to June 2015 blood samples were obtained from 38 women eligible for IVF (IVF-group). Women were recruited at the baseline consultation and then they were followed until pregnancy test was performed. Biochemical diagnosis of pregnancy was the end-point of the study. According to the levels of β-chorionic gonadotropin (βHCG) women were stratified into two subgroups: IVF-positive if βHCG levels were higher than 100 IU/L and IVF-negative if (βHCG) levels were equal or lower than 100 IU/L. The evolution of the pregnancy was not recorded.
The control group consisted of 30 healthy, age-matched women, without any known hereditary or acquired thrombophilic alteration or personal history of thrombotic or bleeding disorder who had undergone at least one uneventful physically conceived pregnancy and without any personal history of miscarriage. The protocol of the study was in accordance with the commitment of the Helsinki declaration and was approved by the institutional ethics committee. All subjects provided informed written consent before inclusion in the study.
Inclusion criteria
Women were eligible for IVF according to established selection criteria applied in our institution. All women had full blood count, platelet count, prothrombin time, activated partial thromboplastin time, fibrinogen, renal and liver function within the normal range.
Exclusion criteria
Women younger than 18 years or older than 45 years, weight less than 50 kg or more than 100 kg, with a personal or family history of venous thromboembolism (VTE) or hemorrhagic syndromes, known hereditary or acquired thrombophilia, active anticoagulant or antiplatelet treatment or use of these agents during the last 30 days before inclusion, hospitalization for any reason within the previous 3 months, abnormal full blood count or platelet count and ongoing cardiovascular, renal or liver disease, malignancy, or arterial hypertension, known systematic or chronic disease (autoimmune syndrome, heart disease, severe or uncontrolled thyroid disease or HIV infection), treatment with non-steroid anti-inflammatory drugs within the last 10 days before inclusion, ovarian insufficiency (FSH > 9 IU/ml and/or number of antral follicles <8) or polycystic ovary syndrome (defined according to the Rotterdam criteria).
Hormone treatment for the artificial reproductive technique
Estrogen production was first down-regulated to induce controlled ovarian stimulation. Three different protocols of down regulation were used: a long gonadotropin-releasing hormone (GnRH) agonist or a short agonist or an antagonist. Ovarian stimulation was done with recombinant human follicular stimulating hormone (FSH) at doses ranging from 75 IU to 450 IU per day depending on age, body mass index (BMI), antral follicle count, size and number of follicles and estradiol levels (E2). This stimulation was initiated once pituitary desensitization had been achieved (E2 level <50 pg/mL). The response was followed by E2 measurement six days later and by ultrasound scanning of the ovarian follicles at days 9–10 after the first FSH injection, and repeated when necessary. Transvaginal oocyte retrieval was scheduled 35 to 36 h after human recombinant chorion gonadotrophin (hCG) injection and embryo transfer was performed 2–3 days later. On day 2, individually cultured embryos were evaluated on the basis of the number of blastomeres, blastomere size, fragmentation rate and presence of multinucleated blastomeres. Therefore, the ovocytes were retrieved 10 to 14 days after starting the stimulation with the FSH.
Outcomes
Achievement of biochemical pregnancy was the outcome of the study. Pregnancy was controlled by quantitative measurement of βHCG 15 days after embryo transfer.
Blood sampling
Blood samples were collected before the administration of any hormone treatment (T0) and during the IVF procedure as follows: at the maximal down-regulation of the menstrual cycle; between the 5th and 8th day from the administration of the GnRH agonist (T1); at maximal stimulation after treatment by FSH and before hCG injection (T2) and two weeks after gonadotropin-releasing hormone (GnRH) injection (T3). Blood samples were obtained by atraumatic puncture of the antecubital vein, using a 20-gauge needle without tourniquet, into siliconized vacutainer tubes containing 0.105 mol/L trisodium citrate; 1/9 v/v (Becton and Dickinson, France). Platelet-poor plasma (PPP) was obtained by double centrifugation at 2000 g for 20 min at room temperature and plasma aliquots were stored at −80 °C until assayed. Samples were assessed within two weeks after collection. Thromboelastometry was carried out with fresh whole blood.
Molecular and functional analysis
Thrombin generation assay
. Thrombin generation in plasma was assessed using the Calibrated Automated Thrombogram assay (CAT®, Diagnostica Stago, Asnières France) according to manufacturers’ instructions, in the presence of optimal concentrations of TF (5 pM) and procoagulant phospholipids (4 μM) using the PPP-Reagent®. Assay’s performance has been published elsewhere [
12,
17].
Minimal TF-triggered whole blood thromboelastometry (min TF-WB TEM) was assessed in citrated whole blood, on the ROTEM® instrument (TEM®, Munich, Germany). Thromboelastometry was performed at 37 °C in citrated fresh whole blood within 30 min after veinipuncture using 5 pM of TF as described elsewhere [
18]. The following parameters of the thromboelastometric trace were analyzed: (a)
Clotting time (CT, in sec): time from the start of the sample run to the point of first significant clot appearance corresponding to an amplitude of 2 mm, (b)
Clot formation time (CFT, in sec): time from CT until the level of clot firmness reaches an arbitrary value of 20 mm, (c)
α-angle (degree): measurement of clot development kinetics, (d)
Maximum clot firmness (MCF in mm): the maximum vertical amplitude of the thromboelastogram.
Procoagulant phospholipid-dependent clotting time (Proag-PPL) was measured with STA®Procoag-PPL, (Diagnostica Stago, Asnières, France) according to the manufacturer’s instructions as described elsewhere. The inter- and intra-assay coefficients of variation were 3 and 4% respectively.
Thrombomodulin activity. Plasma levels of thrombomodulin activity (TMa) were measured with a functional test on the STA-R analyzer (Diagnostica Stago, Asnières, France) as described elsewhere [
19]. The inter- and intra-assay coefficients of variation were 5 and 6% respectively.
Specific TF activity. Tissue Factor activity (TFa) in PPP was measured with a clotting-based assay as previously described [
20,
21]. The inter- and intra-assay coefficients of variation were 7% and 5% respectively. The levels of, FVIII, FvW, D-Dimers and fibrinogen were measured with conventional assays according to the manufacturer’s instructions (Diagnostica Stago, Asnières, France).
Statistical analysis
The calculation of the sample size was based on the minimum number of patients required for a significant power for the detection of differences (a) between the IVF and control group, (b) at the IVF at the studied time points, (c) between IVF positive and IVF negative groups. The minimum sample size of 27 individuals for each group (IVF and control as well as IVF at each time point) was defined to warrant a two-tail significance at the limit of 5% and a prediction power of 95% with a two-sided α level of 0.05. Regarding the sub-group analysis (IVF positive and IVF negative) the minimum size of 15 patients for each group warrants two-tail significance at the limit of 5% and a prediction power of 85% with a two-sided α level of 0.05.
Special effort and attention was given to avoid missing values. The data are presented as mean ± sd. The Mann–Whitney test for independent samples was used for the comparisons of the studied parameters between the IVF and the control group and between the IVF-positive and IVF-negative group. Non-parametric Wilcoxon test for related samples and ANOVA test were applied to compare changes in variables at the studied time points during the observation period. Pearsons’ test was applied to control correlation between thrombogram parameters and studied blood coagulation variables. Dichotomous variables were compared with χ
2 test. The Upper Normal Limit (UNL) and the Lower Normal Limit (LNL) for each parameter of the studied variables were defined in the control group as follows: UNL = mean + 2 SD, and LNL: = mean – 2 SD. The UNL and LNL of the studied biomarkers were defined in the control group and were compared to the corresponding normal reference range used by our laboratory. The normal ranges have been established according to the requirements for the good quality of laboratory practice by performing the tests in healthy individuals representative of the general population regarding age, sex, ethnicity, BMI. Two-sided values of p < 0.05 were considered as statistically significant.
The model development started by defining the positive diagnosis of pregnancy (if βHCG levels were equal or lower than 100 IU/L) as the dependent variable. The first step consisted of the univariate analysis in order to identify the variables associated with positive pregnancy. The selection of independent variables (which are the biomarkers of hypercoagulability) was done at the level of 5% using the stepwise procedure. The multivariable linear regression model was used to explore the effect of the independent variables on pregnancy outcome. The variables, found to be significant in the univariate analysis (p < 0.05) were included in the multivariate analysis. The variable with the highest p value was excluded from the model. The discrimination capacity of the model was tested with receiver operating characteristics (ROC) analysis and the area under the curve (AUC) was calculated for the quantification of the discrimination capacity of the model. The SPSS statistical software package (Chicago inc 6.1) was used for statistical analysis. Values are mean ± standard deviation.
Discussion
The present prospective longitudinal observational study demonstrates that at baseline, women eligible for IVF present blood hypercoagulability which is characterized by significant increase of platelet and endothelial cell activation biomarkers. The baseline state of cellular hypercoagulability, which persisted practically unchanged during the period of hormone treatment administration for IVF, was consisted of significantly shortening of Procoag-PPL clotting time. This test is correlated with increased concentration of procoagulant microparticles derived from platelets or other cells [
22,
23]. The levels of TFa and TMa, at baseline and during hormone treatment, were also significantly increased as compared to age-matched women with naturally occurring uneventful pregnancies. The TFa and TMa are biomarkers of endothelial cell or platelet activation [
24‐
26]. In addition, increased TFa levels in plasma is a marker of monocyte activation [
27]. The Procoag-PPL, TFa and TMa levels were not significantly modified during treatment for estrogen down-regulation and ovarian stimulation indicating that the shift towards cellular hypercoagulability observed at baseline was not correlated with hormone variations. The data from the present study are in accordance with those recently published by Olausson et al. which demonstrate that platelet, endothelial and monocyte-derived microparticles and inflammation biomarkers are significantly increased in women undergoing IVF [
28]. Thrombin generation was also significantly enhanced but the baseline levels of FVIII, FvW, D-Dimers and fibrinogen were similar to those observed in the control group, confirming previous studies [
4]. To the best of our knowledge, this is the first study showing that infertility is linked to a systemic cell activation that offers procoagulant substances, mainly endothelial cells and platelets. This concept is supported by recent studies reporting that increased levels of plasminogen activator inhibitor 1 (PAI-1), thrombin activatable fibrinolysis inhibitor (TAFI) or tissue factor pathway inhibitor (TFPI), which are synthesized and secreted by activated endothelial cells are related with infertility and IVF failure [
29‐
32].
Furthermore, the present study showed that short Procoag-PPL clotting time as well as increased TFa and TMa levels are independent risk factors for IVF failure. The implication of high levels of TF and procoagulant microparticles in the pathogenesis of infertility, IVF failure and vascular complications during pregnancy, such as recurrent first trimester miscarriage, fetal loss, stillbirth, early and severe pre-eclampsia or prematurity has already been reported by others (reviewed in [
33,
34]). Herein, we demonstrate for a first time that the shortened Procoag-PPL associated with the mean rate index (MRI) of the propagation phase of thrombin generation assessed at the maximal down-regulation of the menstrual cycle (between the 5
th and 8
th day from the administration of the GnRH agonist) are predictors of IVF outcome. These tests could be used in the construction of a risk assessment model for IVF issue. The design of the present study does not allow the identification of the underlying causes that lead to cellular activation and hypercoagulability which is observed at baseline. This investigation appears to be attractive for the elucidation of the link with hypercoagulability in women eligible for IVF. The association between hypercoagulablity and negative IVF outcome has recently been reported by Di Nisio et al., who proposed that the increase of D-Dimers levels in plasma is a predictor for IVF failure [
35]. The implication of increased D-Dimers in the sterility is further supported by the data presented by Di Micco et al. [
36]. The univariate analysis of the data reported herein confirms that the increase of D-Dimers during hormone treatment is a negative prognostic factor for IVF outcome. However, the impact of D-Dimers disappeared in the multivariate analysis indicating that high concentrations of procoagulant phospholipids, detected by the short Procoag-PPL, have a dominant role in negative IVF outcome.
Interestingly, increased thrombin generation was found to be a positive predictor for IVF outcome. This finding is in accordance with the concept that thrombin generation is necessary for blastocyst implantation, remodeling of decidualized human endometrial stromal cells and subsequent trophoblast invasion and remodeling of the spiral arteries and arterioles; a process driven by TF expression in the endometrial microenvironment which is in contact with mother’s blood [
7‐
10]. The concept that the non-suffering of endothelial cells, platelets or other cells that potentially release procoagulant phospholipids, is determinant for a positive IVF outcome is supported by our study. Indeed, the presence of high levels of procoagulant phospholipids in plasma, as expressed by shortened Procoag-PPL was the dominant parameter related with IVF failure while the decrease of thrombin generation was found to be a complementary factor. Control measurements showed that factor V and factor II levels were within the normal range (data not shown) ruling out clotting factors’ consumption. The reasons for which thrombin generation is decreased in women with IVF failure have to be investigated.
At all studied time points the parameters of thromboelastometry were not significantly different between the IVF-group and the control group. Fibrinogen levels, platelet count and hematocrit are variables with a major impact on thromboelastometric profile. Both IVF-group and the control group had these variables within the normal ranges.
We found that when the ovarian stimulation was maximal, thrombin generation, FVIII, FvW and D-Dimers levels were significantly increased, in agreement with older studies [
5,
37‐
41]. However, these changes were not reflected on the kinetics of clot formation and its qualitative characteristics when coagulation was triggered by a low TF concentration. In women undergoing hormone treatment for IVF preparation thromboelastographic analysis performed after triggering contact system showed a slight but significant acceleration of the kinetics of clot formation following GnRH treatment indicating that peak concentrations of estrogens are associated with a possible enhancement of FXII activation [
42]. Women who had baseline thrombin generation above the UNL remained at the same levels during the down-regulation and the stimulation phase of the treatment (data not shown). Therefore, our data, in agreement with previous studies [
43] and support the concept that hormone treatment for IVF represents a mild procoagulant stimulus, which has a minor effect on the global haemostatic balance, the kinetics of clot formation or its qualitative characteristics. Two weeks after hCG injection and embryo transfer, thrombin generation, TMa, TFa and Procoag-PPL remained without any significant modifications as compared to the phase of ovarian hyperstimulation. In contrast, the D-Dimers tended to increase and this finding is in agreement with others [
2]. De Nisio et al., found that one week after the administration of gonadotropin, D-Dimers levels increased considerably [
35]. We also found that even two weeks after GnRH injection D-Dimers levels are still increased, reflecting enhanced fibrin formation or fibrinolysis following r-hCG, as described by Biron et al. [
2].
The present study has some limitations. The number of the enrolled women, although it provides sufficient statistical power to identify the most relevant biomarkers related with IVF outcome, does not allow generalizability of the model. The group of IVF women was heterogeneous regarding the IVF indication. In some women the indication of the IVF was of male origin. Although the elimination of these cases did not significantly influence the final results, this does not warrant that the presence of this subpopulation had no impact on the results. A new prospective study in a larger population is planned to corroborate these findings. Women enrolled in the study were treated with three different protocols for estrogen down-regulation. The subgroup analysis according to the therapeutic protocol did not demonstrate any significant difference on the studied biomarkers among the subgroups (data not shown). Several different hormone treatments and protocols are applied for IVF preparation by the different IVF centers. Whether this protocol variability influences the kinetics of the studied biomarkers has to be investigated. However, the major findings of our study, which are (a) the presence of cell derived hypercoagulable state at the baseline, before any treatment administration and (b) the predictive value of the Procoag-PPL clotting time associated with the MRI of thrombin generation on IVF outcome are not influenced by the subsequent hormone treatments for ovarian stimulation. Although this study fulfils the criteria for the statistical power in the selection of the most clinically relevant biomarkers, a new prospective independent validation of this model is required in a larger multicenter study.
It has been suggested that heparin may improve the intrauterine environment in sub-fertile women, by enhancing growth factors to improve attachment of the embryo to the lining of the womb [
44]. Thus low molecular weight heparin (LMWH) is often offered to women eligible for IVF as an adjunct treatment in an attempt to improve the probability for a positive outcome. However, a recent meta-analysis showed that it is unclear whether peri-implantation heparin administration in assisted reproduction treatment cycles improves clinical pregnancy rates in sub-fertile women [
45]. The elaboration of a risk assessment model with the clinically relevant biomarkers of hypercoagulability proposed by the present study is of clinical interest in order to identify women who could benefit from the peri-implantation administration of antithrombotic treatment.
Acknowledgements
The authors acknowledge the nurses from the Department of Obstetrics and Gynecology for their efforts to collect blood samples. The authors wish to thank the skilful technical assistance of the technicians of the Laboratory of Haemostasis, Service d’Hématologie Biologique, Hôpital Tenon and particularly Mme Marie Paule Roman and Severine Bouffard.