Unexplained pregnancy loss: a marker of basal endothelial dysfunction?

ORIGINAL ARTICLES: EARLY PREGNANCY

Unexplained pregnancy loss: a marker of basal endothelial dysfunction? Elisabeth Pasquier, Ph.D.,a Luc De Saint Martin, Ph.D.,a Caroline Bohec, M.D.,b Michel Collet, Ph.D.,b Franc¸oise Dignat George, Ph.D.,c and Dominique Mottier, Ph.D.a a EA 3878 (GETBO), Department of Internal Medicine and Chest Diseases, Brest University Hospital, La Cavale Blanche Hospital, Brest; b Division of Gynaecology, Brest University Hospital, Morvan Hospital, Brest; c UMR-S608 INSERM,  de Pharmacie, Universite  de la Me diterrane e, Marseille, France F-Marseille, and Faculte

Objective: To compare the microparticle levels of women referred for unexplained pregnancy loss with those of parous controls. Design: Incident case-control study. Setting: University medical center. Patient(s): 124 women consecutively referred for unexplained pregnancy losses (two or more losses at or before 21 weeks of gestational age, or at least one later loss), and 273 parous women without pregnancy loss. Intervention(s): Numeration of circulating microparticles by flow cytometry after differentiation of subpopulations according to the expression of membrane-specific antigens (CD51, CD144, or CD146 for endothelial, CD41 for platelet, CD45 and CD66b for leukocyte and neutrophil microparticles). Main Outcome Measure(s): Plasma levels of microparticles. Results: A relative hypercoagulable state assessed by thrombin generation test had been previously reported in such cases, so we hypothesized that this could be explained by an excess of procoagulant microparticles. The study women displayed statistically significantly lower platelet and higher endothelial microparticle levels than the controls. The parameters of the thrombin generation test were only correlated with the level of endothelial microparticles, with a low coefficient of Speerman’s correlation (r¼0.15). Conclusion(s): The difference in microparticle levels between the patients and controls does not clearly explain the hypercoagulable state reported in the patients but could reflect chronic endothelium damage. (Fertil SterilÒ Use your smartphone 2013;100:1013–7. Ó2013 by American Society for Reproductive Medicine.) to scan this QR code Key Words: Case-control study, endothelium, microparticles, pregnancy loss, thrombin and connect to the generation test Discuss: You can discuss this article with its authors and with other ASRM members at http:// fertstertforum.com/pasquiere-pregnancy-loss-endothelium/

P

revious investigations of the role of genetic thrombophilic markers in unexplained pregnancy loss have yielded conflicting results (1). In a large case-control study, we found no association between unexplained pregnancy loss (early, recurrent, or late) and factor V Leiden (FVL) or prothrombin G20210A (PTG) mutations in either parent (2). To investigate

the possibility of undetectable thrombophilia and interactions between different prothrombotic abnormalities, we used a global test of the hemostatic pathway in the same group enrolled in our previous study (3). Using the method described by Hemker et al. (4), we explored the potential of thrombin generation in these nonpregnant women. Especially when the pregnancy

Received February 6, 2013; revised May 30, 2013; accepted June 4, 2013; published online July 2, 2013. E.P. has nothing to disclose. L.D.S.M. has nothing to disclose. C.B. has nothing to disclose. M.C. has nothing to disclose. F.D.G. has nothing to disclose. D.M. has nothing to disclose. The study was supported by a grant (PHRC 2002) from the French Health Ministry. Reprint requests: Elisabeth Pasquier, Ph.D., EA 3878 (GETBO), Department of Internal Medicine and Chest Diseases, Brest University Hospital, La Cavale Blanche Hospital, 29609 Brest, France (E-mail: [email protected]). Fertility and Sterility® Vol. 100, No. 4, October 2013 0015-0282/$36.00 Copyright ©2013 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2013.06.008 VOL. 100 NO. 4 / OCTOBER 2013

discussion forum for this article now.*

* Download a free QR code scanner by searching for “QR scanner” in your smartphone’s app store or app marketplace.

loss had occurred between weeks 9 and 12 of gestational age, the patients displayed a baseline hypercoagulable state of relative thrombomodulin resistance that was not due to either factor VIII or protein S plasma levels or to FVL or PTG mutations (3). Circulating microparticles (MPs) are submicron vesicles released from cell membranes in response to activation or apoptosis. As both markers and effectors of the ongoing process, MPs increase in a variety of disease states, such as thrombosis and vascular dysfunction. In addition to their many other properties, MPs are well known for their procoagulant activity, which mainly depends on phosphatidylserine and tissue factor expression. We 1013

ORIGINAL ARTICLE: EARLY PREGNANCY hypothesized that the hypercoagulable state assessed in our patients at least 2 months after any recognized obstetric event could be explained by an excess of procoagulant MPs, as reported by Laude et al. (5). Therefore, we explored MPs originating from platelets (PMP), endothelium (EMP), leukocytes (LMP), and neutrophils (NMP) in the first 124 women referred for a history of unexplained pregnancy loss and in 273 controls. Both groups had been previously enrolled in a study of thrombophilic mutations (2) and assessment of the potential to generate thrombin (3).

Controls

MATERIALS AND METHODS

Samples

Study Design and Setting

For the MP numeration (6), we obtained venous blood samples by venipuncture with 0.129 mol/L sodium citrate tubes. Blood samples were centrifuged at 1,500  g for 15 minutes within 2 hours of venipuncture. The plasma was then recovered and centrifuged at 13,000  g for 2 minutes.

Our incident case-control study compared the MP plasma levels of women referred for unexplained pregnancy loss with those of parous controls who were at least 2 months from any recognized obstetric event. The women were recruited from February 2003 to March 2008 at the University Hospital of Brest. They were seen once by an investigator for the inclusion visit, during which they underwent a medical review using a standard questionnaire and a venous blood puncture. The blood samples were always collected randomly throughout the menstrual cycle but at least 2 months after any recognized obstetric event as well as after any anticoagulation, antiplatelet, or estrogen/progesterone treatment.

Patients All women enrolled in the study were aged between 18 and 45 years, were from the West Brittany area, and had been consecutively referred for a history of unexplained pregnancy loss by obstetricians in the area either in private and/or public practice who participate in our reliable obstetric network (2). Pregnancy loss was defined as two or more unexplained consecutive miscarriages at or before 21 weeks of gestational age (EPL group), or at least one unexplained stillbirth after 21 weeks (LPL group). The exclusion criteria consisted of maternal or paternal carrier of a structural chromosomal rearrangement, maternal persistent antiphospholipid antibodies, or any anatomic abnormality likely to be responsible for pregnancy loss. A standard evaluation comprised maternal testing for lupus anticoagulant, determination of immunoglobulin G and M antibodies against cardiolipin and b2-glycoprotein-1, a hysterography or hysteroscopy, and maternal and paternal cytogenetic analysis. Additionally concerning late pregnancy losses, any recognized (maternal or fetal) cause of fetal loss, comprising preeclampsia and placental abruption, was considered as an exclusion criterion; a systematic histopathologic examination of the fetus and the placenta was performed to improve the detection of infection, placental abruption, placenta membranacea, or circumvallate placenta, which are among the well-known causes of fetal loss. It is worth noting that a premature rupture of membranes before 20 weeks of gestation and a history suggesting cervical incompetency or fetal infection were considered to be exclusion criteria. 1014

The controls were enrolled from the same geographical area, among women aged between 18 and 45 years, using the electoral registers. Women were potentially eligible if they had given birth to at least one living child. Exclusion criteria included previous pregnancy loss and preclinical miscarriages (2, 3). For both patients and controls, a history of venous thromboembolism (idiopathic or not), the use of assisted reproduction, or polycystic ovary syndrome were not considered to be exclusion criteria.

Reagents Biocytex supplied the phycoerythrin (PE) conjugated monoclonal antibodies against CD146. Beckman-Coulter Immunotech provided the fluorescein isothiocyanate (FITC) conjugated annexin V to label the total MPs, the PEconjugated and FITC-conjugated isotype controls to define the background noise of the labeling, the PE-conjugated monoclonal antibodies against CD144, and the FITCconjugated monoclonal antibodies against CD45, CD66b, and CD51 to label the MP subpopulations.

Numeration of Circulating Microparticles by Flow Cytometry The membrane MP subpopulations were differentiated in platelet-free plasma according to the expression of membrane-specific antigens. Numeration of EMP was performed using anti-CD51, anti-CD144, or anti-CD146 labeling; numeration of PMP, LMP, and NMP was performed using anti-CD41, anti-CD45, and anti-CD66b labeling, respectively. Moreover, regardless of their cellular origin, annexin V binding was used to numerate the phosphatidylserine-expressing circulating MPs. After labeling and dilution, samples were analyzed by flow cytometry on an EPICS XL (Beckman Coulter). The MPs present in plasma were analyzed according to their parameters of size and fluorescence. Using 0.8-mm latex beads, we defined MP as vesicles <1 mm in diameter and positively them labeled with specific monoclonal antibodies (compared with IgG isotype-matched controls) or annexin V. The absolute values of MPs were calculated using Flowcount beads and expressed as MPs per microliter (mL) of plasma.

Thrombin Generation Test Thrombin generation was performed in the whole plasma according to the method described by Hemker et al. (4) in a Fluoroscan Ascent fluorometer (Thermo Labsystems OY) equipped with a dispenser, as previously reported elsewhere (3). The analyzed parameter was the endogenous thrombin VOL. 100 NO. 4 / OCTOBER 2013

Fertility and Sterility®

TABLE 1 Main characteristics of patients and controls. Characteristics

Controls (n [ 273)

Mean age, y (SD) 35.05 (4.4) 22.7 (4.4) BMI, kg/cm2 (SD) Tobacco use, n (%) 77 (28.2) Venous thromboembolism (%) 7 (2.6) Free protein S, % (SD) 85.3 (17.8) Factor VIII, % (SD) 101.3 (28.5) 12 (4.4) Factor V Leiden, n (%)a 6 (2.2) Prothrombin G20210A, n (%)a ETP without TM, mM  mn (SD) 2.13 (0.30) ETP with TM, mM  mn (SD) 1.16 (0.39) Inhibition-r, % (SD) 45.7 (16.1)

Cases (n [ 124)

P value

33 (4.4) < .001 23.2 (5.2) .34 31 (25) .54 4 (3.2) .47 88.9 (20.0) .077 95.6 (24.2) .055 7 (5.6) .62 3 (2.4) 1.00 2.11 (0.29) .41 1.3 (0.35) < .001 37.9 (12.3) < .001

Note: BMI ¼ body mass index; ETP ¼ endogenous thrombin potential; inhibition-r ¼ TM related inhibition rate of the ETP; SD ¼ standard deviation; TM ¼ thrombomodulin. a No patients or controls carried homozygous mutation or combined mutations. Pasquier. Microparticles and pregnancy loss. Fertil Steril 2013.

potential (ETP) corresponding to the area under the curve, measured in the absence or presence of thrombomodulin. The effect of thrombomodulin on thrombin generation was expressed as the thrombomodulin-related inhibition of the ETP (inhibition-r), defined as the ratio of the absolute difference in ETP, determined in the absence or presence of thrombomodulin, to the ETP in the absence of thrombomodulin.

Statistical Analysis Statistical analysis was performed using SPSS Statistics v.19. Student’s t-test and chi-square test were used for normal parameters and categorical variables, respectively. Because the MP levels were not normally distributed, the results were expressed as median and extreme values. The MP levels were compared between groups using the MannWhitney U test. In addition, the distribution of MP levels in the patients and controls was compared using the chisquare test, according to the quartiles of the control distribution. This approach permits calculation of the odds ratio (OR) with 95% confidence interval (CI) as an estimate of the risk of pregnancy loss associated with MP levels. Using logistic regression, an adjustment for both predefined variables (pro-

tein S, factor VIII, FVL, and PTG) and potential confounding variables (P< .2) was performed. The predefined variables were selected according to our previous results (2) and the literature. Additionally, an adjustment for the elapsed time since the last obstetric event (considered as a continuous variable) was planned. Indeed, this is an implicit bias of the study design (median of 133 days, min–max: 60–3,165 for patients; median of 1,465 days, min–max: 69–6,507 for controls). Associations between cell-derived MP levels and thrombin generation parameters were assessed by using Spearman’s correlation. A subanalysis according to the mean gestational age at pregnancy loss and the number of losses was performed. The study was approved by the local ethics committee, the CPP of Brest University Hospital, and was supported by a grant (PHRC) from the French Health Ministry. The authors have no conflicts of interest in the research.

RESULTS The MP measurements were available for the first 124 patients and 273 controls previously enrolled (2, 3). Most of the differences in population characteristics between groups were similar to our previously reported ones in 2009 and 2011 (2, 3) (Table 1). The body mass index, tobacco use, rate of venous thromboembolic events, FVL/PTG distribution, and both factor VIII and protein S levels did not differ between the groups. However, the patients were statistically significantly younger than the controls (33  4.4 vs. 35.05  4.4 years; P< .001). In the EPL group (median of three losses, min–max: 2–16), 71 women had experienced three or more early losses, and 8 an additional late loss (after week 21). In the presence of thrombomodulin, the ETP was statistically significantly enhanced in the patients compared with the controls (1.3  0.35 vs. 1.16  0.39 mM , mn; P< .001) whereas there was no statistically significant difference in the absence of thrombomodulin. Thus, the thrombomodulin-related inhibition of the ETP (inhibition-r) was 37.9  12.3% vs. 45.7  16.1% for patients and controls, respectively (OR 0.031; 95% CI, 0.007–0.14). After adjustment for age, protein S, factor VIII, FVL, PTG, and elapsed time since the last obstetric event, these results remained similar (OR 0.007; 95% CI, 0.001–0.068).

TABLE 2 Distribution of microparticle counts (n/mL) in the groups. Controls (n [ 273) MP count (n/mL) in each group PMP CD51þ EMP CD146þ EMP CD144þ EMP CD45þ MP CD66bþ MP

All cases (n [ 124)

EPL group (n [ 107)

LPL group (n [ 25)

Median

Min-max

Median

Min-max

Median

Min-max

Median

Min-max

454 8 3 6 14 10

(35–5,983) (1–43) (0–36) (1–45) (1–208) (1–120)

319 8 5 10 20 11

(48–5,519) (0–69) (0–61) (0–56) (1–108) (0–54)

315 8 5 10 19 11

(48–5,519) (0–69) (0–61) (0–56) (1–108) (0–54)

355 7 7 10 22 15

(75–1,236) (1–39) (1–39) (1–54) (1–54) (1–52)

Note: EMP ¼ endothelial microparticle; EPL group ¼ two or more unexplained consecutive miscarriages at or before 21 weeks of gestational age; LPL group ¼ at least one unexplained pregnancy loss after 21 weeks; MP ¼ microparticle; PMP ¼ platelet microparticle. Pasquier. Microparticles and pregnancy loss. Fertil Steril 2013.

VOL. 100 NO. 4 / OCTOBER 2013

1015

ORIGINAL ARTICLE: EARLY PREGNANCY

TABLE 3 Subgroup analysis using the Mann-Whitney U test, comparison of controls with patients after stratification according to the mean gestational age at miscarriage and the number of losses. Comparison with controls, P value Subgroup All cases (n ¼ 124) EPL (n ¼ 107) R3 EPL (n ¼ 71) ¼2 EPL (n ¼ 36) LPL (n ¼ 25) MGT %9 (n ¼ 67) %7 (n ¼ 33) 9 < MGT % 12 (n ¼ 27)

PMP

CD51D EMP

CD146D EMP

CD144D EMP

CD45D MP

CD66bD MP

< .001 < .001 .001 .023 .063

.27 .25 .24 .73 .75

.008 .03 .006 .92 .023

.006 .02 .035 .19 .016

.058 .10 .008 .39 .17

.11 .223 .011 .15 .25

< .001 < .001 .008

.2 .97 .5

.023 .11 .01

.2 .14 .3

.133 .98 .03

.330 .84 .13

Note: EMP ¼ endothelial microparticle; EPL group ¼ two or more unexplained consecutive miscarriages at or before 21 weeks of gestational age; LPL group ¼ at least one unexplained pregnancy loss after 21 weeks; MGT ¼ mean gestation time; MP ¼ microparticle; PMP ¼ platelet microparticle. Pasquier. Microparticles and pregnancy loss. Fertil Steril 2013.

As the PMP and annexin Vþ MP levels were highly correlated (r¼0.97; P< .001), we did not report the results of annexin Vþ MPs. We found statistically significantly lower PMP levels in the patients compared with the controls (319 [48–5519] vs. 454 [35–5983] counts/mL; P< .001). In addition, we found statistically significantly higher CD146þ and CD144þ EMP levels in the patients compared with the controls (CD146þ: 5 [0–61] vs. 3 [0–36] counts/mL, P¼ .008; CD144þ: 10 [0–56] vs. 6 [1– 45] counts/ml, P¼ .006). The CD51þ EMP, CD66bþ and CD45þ MP levels were not different between the patients and the controls (Tables 2 and 3). Adjusting for age, protein S, factor VIII, FVL, PTG, and elapsed time since the last obstetric event did not influence our results. The risk of pregnancy loss according to different MP level cutoffs is shown in Table 4. The ORs are around 0.5 for increased PMP (cutoff <50%: OR 0.46; 95% CI, 0.29–0.71) and around 2 for increased CD144þ (cutoff <50%: OR 2.21; 95% CI, 1.39–3.52) or CD146þ (cutoff <75%: OR 2.05; CI, 1.30–3.20) EMP levels. We studied the association between MP levels and thrombin-generation parameters using Spearman’s correlation. We found only that the thrombomodulin-related inhibi-

tion of the ETP (inhibition-r) was statistically significantly and negatively correlated with the CD146þ and CD144þ EMP levels (r¼0.15 and P¼ .03) in the patients. After adjustment for EMP levels, the differences in ETP and inhibition-r observed between the patients and the controls remained highly statistically significant. It should be pointed out that we detected no statistically significant correlation between PMP levels and thrombin generation parameters. We also conducted a subanalysis according to the gestational age at pregnancy loss (Table 3). Despite the small sample size, the CD146þ and CD144þ EMP levels in the women with late pregnancy loss (after week 21) remained statistically significantly higher than in controls (Table 2) whereas the PMP levels were especially low in women with very early loss (before week 7) (P< .001). Finally, we conducted a subanalysis according to the number of losses in the EPL group. In the subgroup of patients with three or more losses, the CD45þ and CD66bþ MP levels were also statistically significantly higher than in controls. In the patients with two early losses, only PMP levels were statistically significantly different compared with controls (see Table 3).

TABLE 4 Risk of pregnancy loss associated with increasing percentiles of microparticle levels. Cutoff (%)

PMP

CD51D EMP

CD146D EMP

CD144D EMP

CD45D MP

CD66bD MP

0.52 (0.33–0.83) 0.5 (0.3–0.9)a

1.16 (0.67–2.03) 1.45 (0.7–3)a

0.45 (0.03–7.30) –b

1.36 (0.81–2.26) 1 (0.58–1.9)a

1.14 (0.67–1.94) 1.67 (0.8–3.45)a

1.17 (0.69–2.01) 1.72 (0.83–3.58)a

0.46 (0.29–0.71) 0.42 (0.24–0.73)a

1.04 (0.68–1.60) 1.13 (0.7–1.9)a

1.56 (1.01–2.42) 1.98 (1.13–3.45)a

2.21 (1.39–3.52) 2.36 (1.31–4.24)a

1.84 (1.18–2.86) 2.77 (1.54–4.99)a

1.43 (0.93–2.22) 2 (1.16–3.52)a

0.51 (0.29–0.9) 0.44 (0.2–0.93)a

1.22 (0.76–1.96) 1.38 (0.8–2.4)a

2.05 (1.30–3.20) 2.32 (1.35–3.99)a

1.64 (1.01–2.7) 1.76 (0.95–3.2)a

1.34 (0.84–2.13) 1.49 (0.8–2.6)a

1.51 (0.95–2.4) 1.75 (1–3.05)a

<25 <50 <75

Note: Results are shown as the odds ratio (95% confidence interval). The cutoff levels are set at quartiles of the control group distribution. The comparison of the patients (n ¼ 124) with the control group (n ¼ 273) was performed using the chi-square test. EMP ¼ endothelial microparticles; MP ¼ microparticle; PMP ¼ platelet microparticle. a Comparison of the controls with the patient subgroup (n ¼ 71) defined by three or more losses before week 21. b No patients in the first quartile of the control distribution. Pasquier. Microparticles and pregnancy loss. Fertil Steril 2013.

1016

VOL. 100 NO. 4 / OCTOBER 2013

Fertility and Sterility®

DISCUSSION At least 2 months after any obstetric event, we found statistically significantly lower PMP and higher EMP levels in women with a history of unexplained pregnancy loss compared with healthy parous controls. Normal pregnancy is characterized by increased levels of MPs, but the prevalence, cell origin, and role of MPs in pregnancy-related complications remain controversial (7). Several studies have stated that circulating MPs some time after pregnancy, quantified using a thrombin generation test or measured by flow cytometer (for PMPs or EMPs), may be increased in women with past history of fetal loss (5, 8, 9). However, others have failed to demonstrate any such difference in MP levels (10). The main shortcomings of those previous studies were the small sample sizes and the discrepancy between the methods used (11). Decreased PMP levels have been reported during preeclampsia (6) or severe sepsis but never in non pregnant women with past history of pregnancy loss. As hypothesized in these clinical settings, PMP could be cleared from the plasma by adhering to blood or vascular cells at site of endothelial damage in our patients. The enhanced thrombin generation potential reported in our patients (2) could be partly explained by the higher CD144þ and CD146þ EMP levels. However, we believe that other underlying mechanisms are clearly involved, as suggested by our results (i.e., the low coefficient of the Spearman’s correlation, r ¼ 0.15). Tissue factor pathway inhibitor deficiency has a proven effect on thrombin generation and could be a more plausible explanation (12). In addition, the patients with late loss (LPL) had statistically significantly higher EMP levels than the controls (P¼ .023 and .016 for CD146þ and CD144 þ EMP, respectively) (Table 3), whereas the thrombin generation potential was not different between this subgroup of patients and the controls (2). This also suggests that the data from MP numeration and thrombin generation test do not exactly reveal the same biologic dysfunctions in our patients. Thus, the detection of raised EMP (and decreased PMP) levels some time after pregnancy could reflect endothelium damage rather than a prothrombotic state. Incidentally, we had previously reported significantly raised levels of circulating CD146, a known marker of endothelial dysfunction, in the first 100 patients compared with their age-matched controls (13). In the literature, raised EMP levels have been associated with most of the cardiovascular risk factors and have correlated with established parameters of endothelial dysfunction. The EPIC cohort study suggested that recurrent miscarriage and stillbirth are strong sex-specific predictors

VOL. 100 NO. 4 / OCTOBER 2013

for myocardial infarction (14). Their results have been recently confirmed by another cohort study (15). Our data provide an additional argument for basal vascular dysfunction in women with pregnancy loss.

REFERENCES 1.

2.

3.

4.

5.

6.

7. 8.

9.

10.

11.

12.

13.

14.

15.

Kist WJ, Janssen NG, Kalk JJ, Hague WM, Dekker GA, de Vries JI. Thrombophilias and adverse pregnancy outcome—a confounded problem! Thromb Haemost 2008;99:77–85. Pasquier E, Bohec C, Mottier D, Jaffuel S, Mercier B, Ferec C, et al. Inherited thrombophilias and unexplained pregnancy loss: an incident case-control study. J Thromb Haemost 2009;7:306–11. De Saint Martin L, Duchemin J, Bohec C, Courturaud F, Mottier D, Collet M, et al. Increased thrombin generation measured in the presence of thrombomodulin in women with early pregnancy loss. Fertil Steril 2011;95:1813–5. Hemker HC, Giesen P, Al Dieri R, Regnault V, de Smed E, Wagenvoord R, et al. The calibrated automated thrombogram (CAT): a universal routine test for hyper- and hypocoagulability. Pathophysiol Haemost Thromb 2002;32:249–53. Laude I, Rongieres-Bertrand C, Boyer-Neumann C, Wolf M, Mairovitz V, Hugel B, et al. Circulating procoagulant microparticles in women with unexplained pregnancy loss: a new insight. Thromb Haemost 2001;85: 18–21. Bretelle F, Sabatier F, Desprez D, Camoin L, Grunebaum L, Combes V, et al. Circulating microparticles: a marker of procoagulant state in normal pregnancy and pregnancy complicated by preeclampsia or intrauterine growth restriction. Thromb Haemost 2003;89:486–92. Aharon A, Brenner B. Microparticles and pregnancy complications. Thromb Res 2011;127:67–71. Carp H, Dardik R, Lubetsky A, Salomon O, Eskaraev R, Rosenthal E, et al. Prevalence of circulating procoagulant microparticles in women with recurrent miscarriage: a case-controlled study. Hum Reprod 2004; 19:191–5. Kaptan K, Beyan C, Ifran A, Pekel A. Platelet-derived microparticle levels in women with recurrent spontaneous abortion. Int J Gynaecol Obstet 2008; 102:271–4. Toth B, Nieuwland R, Kern M, Rogenhofer N, Berkmans R, Rank A, et al. Systemic changes in haemostatic balance are not associated with increased levels of circulating microparticles in women with recurrent spontaneous abortion. Am J Reprod Immunol 2008;59:159–66. Alijotas-Reig J, Palacio-Garcia C, Llurba E, Vilardell-Tarres M. Cell-derived microparticles and vascular pregnancy complications: a systematic and comprehensive review. Fertil Steril 2013;99:441–9. Gardiner C, Cohen H, Austin SK, Machin SJ, Mackie IJ. Pregnancy loss, tissue factor pathway inhibitor deficiency and resistance to activated protein C. J Thromb Haemost 2006;4:2724–6. Pasquier E, Bardin N, De Saint Martin L, Le Martelot MT, Bohec C, Roche S, et al. The first assessment of soluble CD146 in women with unexplained pregnancy loss: a new insight? Thromb Haemost 2005;94:1280–4. Kharazmi E, Dossus L, Rohrmann S, Kaaks R. Pregnancy loss and risk of cardiovascular disease: a prospective population-based cohort study (EPIC-Heidelberg). Heart 2011;97:49–54. Ranthe MF, Andersen EA, Wohlfarht J, Bundgaard H, Melbye M, Boyd HA. Pregnancy loss and later risk of atherosclerotic disease. Circulation 2013; 127:1775–82.

1017