- Email: [email protected]
Tinzaparin for the treatment of foetal growth retardation: An open-labelled randomized clinical trial
Thrombosis Research 170 (2018) 38–44
Contents lists available at ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Full Length Article
Tinzaparin for the treatment of foetal growth retardation: An open-labelled randomized clinical trial☆
T
⁎
Anette Tarp Hansena,b,c, , Puk Sandagerb,d, Mette Ramsinge, Olav B. Petersenb,d, Jannie D. Salvigd, Svend Juulf, Niels Uldbjergb,d, Anne-Mette Hvasa,b a
Aarhus University Hospital, Centre for Haemophilia and Thrombosis, Department of Clinical Biochemistry, Denmark Aarhus University, Department of Clinical Medicine, Denmark c Aalborg University Hospital, Department of Clinical Biochemistry, Denmark d Aarhus University Hospital, Department of Obstetrics and Gynaecology, Denmark e Randers Regional Hospital, Department of Pathology, Denmark f Section for Epidemiology, Department of Public Health, Aarhus University, Denmark b
A R T I C LE I N FO
A B S T R A C T
Keywords: Birth weight Foetal growth retardation Heparin, low molecular weight Placenta, thrombosis Randomized controlled trial
Objective: Foetal growth retardation (FGR) is a leading cause of perinatal death and long-term harms at survivors. Placental infarction plays a role in FGR, yet, no trials have evaluated whether low molecular weight heparins increase birth weight in ongoing FGR pregnancies. Methods: An open-labelled randomized trial in Denmark during 2011–2016, including singleton pregnant women with FGR (estimated foetal weight < 2.3 percentile) diagnosed before gestational weeks 32. Participants were randomly assigned using sealed, blinded envelopes 1:1 to tinzaparin (4500 IU daily until 37 gestational weeks) or no tinzaparin. The primary outcomes were the observed birthweight relative to the expected for gestational age and gender, and foetal growth rates in the two trial groups evaluated by an intention to treat analysis. Results: We enrolled 53 women. The mean gestational age was 261 days in the tinzaparin group and 246 days in the no treatment group. The mean birth weight was 2229 g in the tinzaparin group compared to 1968 g in the no treatment group. However, the birth weight relative to the expected from gestational age and gender was only 2.5 percentage points higher in the tinzaparin group [−5.1 to 10.0] (p = 0.51). The foetal growth rate during follow-up was 124 g/week in the tinzaparin group and 119 g/week in the no treatment group, a difference of 5 g/week [−19 to 29] (p = 0.67). Two perinatal deaths both occurred in the no treatment group. Conclusion: We found no evidence of a tinzaparin effect on the foetal growth rate or the birth weight after adjustment for gestational age.
1. Introduction Worldwide, foetal growth retardation (FGR) is a leading cause of perinatal death and long-term harms for the survivors [1,2]. Furthermore, also the consequences of prematurity challenge these new-borns, as premature delivery is yet the only treatment option. A cornerstone among the pathophysiologic pathways leading to FGR is placental vasculopathy with placental infarction and impaired blood flow in the utero-placental vessels [3,4]. Therefore, anticoagulant drugs may constitute a principle of treatment and among these, low molecular weight heparins are the drugs of choice during pregnancy because they do not
cross the placenta, have a favourable safety profile for the mother, and a predictable anticoagulant effect [5,6,7]. When used as a prophylactic intervention in pregnant women with a prior FGR pregnancy, low molecular weight heparins seems to increase birth weight in subsequent pregnancies [8,9,10]. Furthermore, low molecular weight heparins decrease the uterine artery resistance in hypertensive pregnancies [5,7,12], a mechanism from which also FGR pregnancies may benefit [13,14,15]. However, identification of women at risk of developing FGR is difficult. Yet, there is no evidence on the effect of low molecular weight heparin treatment for FGR diagnosed in ongoing pregnancies. To
☆
Trial registration. ClinicalTrials.gov (EudraCT no. 2011-000818-20). Funding. The Danish Council for Independent Research (0602-02173B FSS) and Central Region of Denmark. Leo Pharma donated the trial drug. ⁎ Corresponding author at: Aalborg University Hospital, Department of Clinical Biochemistry, Hobrovej 18-22, 9000 Aalborg, Denmark. E-mail address: [email protected] (A.T. Hansen). https://doi.org/10.1016/j.thromres.2018.08.006 Received 19 May 2018; Received in revised form 12 July 2018; Accepted 8 August 2018 Available online 10 August 2018 0049-3848/ © 2018 Elsevier Ltd. All rights reserved.
Thrombosis Research 170 (2018) 38–44
A.T. Hansen et al.
address this gap in knowledge, we conducted a randomized trial testing if the low molecular weight heparin, tinzaparin, improves birth weight in ongoing pregnancies complicated by foetal growth retardation. We hypothesized that low molecular weight heparin increases foetal growth in ongoing pregnancies complicated by FGR.
Table 1 Inclusion and exclusion criteria for the trial. Inclusion criteria 1: Singleton pregnant 2: Able to read and understand the written informed consent 3: Foetal growth retardation; estimated foetal weight at least ≤ −22% of expected weight according to gestational age (≈2.3 percentile) with or without Pathologically increased resistance in the uterine arteries; mean pulsatility index > 1.7 standard deviations after gestational weeks 24. In the initial protocol (before trial conduct was commenced), pathologically increased resistance in the uterine arteries from gestational week 19–20 was a separate inclusion criteria.
2. Methods 2.1. Setting and study design We conducted an open-labelled randomized controlled trial at three obstetric centres in Denmark (Aarhus University Hospital, Regional Hospital Randers, and Regional Hospital Herning). The trial was monitored by the Good Clinical Practice Unit at Aarhus University Hospital, Denmark according to “Guideline for good clinical practice for trials on pharmacological products”, World Health Organization, WHO Technical Report Series no 850, 1995, annex 3. The trial was registered in ClinicalTrials.gov (EudraCT no. 2011-000818-20).
Exclusion criteria 1: Age < 18 years 2: Pre-pregnancy maternal body weight > 90 kg 3: Unable to read and understand the written informed consent 4: Renal insufficiency (plasma creatinine > 150 μ/L) 5: Pre-existing hypertension (systolic/diastolic blood pressure > 140/90 mm Hg) 6: Diabetes type I and II 7: Active inflammatory bowel disease 8: Severe heart disease 9: Abuse of alcohol or any drugs 10: Known bleeding disorder (von Willebrand disease, thrombocytopenia, haemophilia) 11: Ongoing treatment with vitamin K antagonists 12: Known allergy to low-molecular weight heparins 13: Previous heparin induced thrombocytopenia (type II) 14: Serious bleeding within the last month 15: Pre-existing indication for anticoagulant prophylaxis with low-molecular weight heparin in current pregnancy 16: Foetal chromosome anomaly 17: Severe foetal malformations 18: Contraindication for tinzaparin 19: Gestational week > 32 weeks
2.2. Participants In the period from November 2011 to August 2016, we included singleton pregnancies diagnosed with FGR before 32 gestational weeks. FGR was defined as an estimated foetal weight below the 2.3 percentile (−2.0 standard deviations (SD); −22% of expected weight corrected for gestational age) with or without increased uterine artery resistance [16]. Increased uterine artery resistance was defined as a mean pulsatility index > 1.7 after gestational weeks 24 [14,15]. We made use of fixed prophylactic dosage of tinzaparin in women with a pre-pregnancy weight between 50 and 90 kg; 4500 international units tinzaparin daily [17]. Therefore, women with a pre-pregnancy weight above 90 kg were not eligible for the present trial [17]. All women attending the Danish antenatal care programme undergo a first trimester risk calculation including ultrasound scan and biochemical markers with predictive value for Down's syndrome. At this ultrasound scan, the crown-rump length is determined as well as a precise gestational age. All participants had this first trimester scan including crown rump length and nuchal translucency measurements as well as risk calculation according to Foetal Medicine Foundation standards, and estimated due date was in all cases calculated from the crown rump length measurement [18]. All study criteria for inclusion and exclusion of trial candidates are shown in Table 1.
Criteria for exclusion from the trial 1: Development of thrombocytopenia, platelet count < 80 × 109/L 2: Allergic reactions to tinzaparin 3: Bleeding requiring hospital admission 4: Non-adherence to study protocol or withdrawal of written informed consent 6: Indication for treatment with a low-molecular weight heparin in current pregnancy 7: Identification of foetal chromosome anomalies 8: Identification of severe congenital foetal heart disease
identification of heparin-induced thrombocytopenia. Furthermore, we advised the obstetricians to wait at least 12 h from the last tinzaparin injection to performing a caesarean section or spinal or epidural analgesic procedures. The participants attended follow-up visits every second week, where an obstetrician specialized in foetal medicine [18] conducted a transabdominal scan. This scan included an estimate of the foetal weight and consecutive Doppler assessments of the utero-placental flow if considered necessary by the obstetrician according to local foetal-medicine guidelines. The Doppler assessment was evaluated by pulsatility indices for the umbilical, uterine, and cerebral arteries [15,19]. Furthermore, a research technician drew a blood sample and obtained an interview according to a systematic case report form; evaluation of adherence to the trial drug, possible side effects, and exclusion criteria. In Denmark, the ultrasound-based monitoring of foetal weights is performed using Marsal's growth charts expressed as percentage deviations in estimated foetal weights according to a gender-specific and gestational age-specific reference curve [16,20]. Data from all scans were stored at the Danish Foetal Medicine Database from which we obtained the estimated foetal weight deviation according to gestational age, expressed as percentage deviation from expected for gestational age [16,22]. All scans were conducted using GE Voluson E8 or E10 ultrasound scanners using curve-linear abdominal probes (GE, Milwaukee, USA). After delivery, placental pathology examination was performed following standard procedures [23]. The pathologist was blinded to
2.3. Primary outcome The primary outcomes were the observed birthweight relative to the expected for gestational age and gender, and foetal growth rates in the two trial groups. Secondary maternal and foetal outcomes are presented in Table 3. 2.4. Trial drug and randomization procedure Upon inclusion, we randomized participants in a 1:1 ratio to either tinzaparin (Innohep®, Leo Pharma, subcutaneous self-injection of 4500 international units daily until completed 37 weeks of gestation) or no tinzaparin (standard procedure), using sealed and blinded randomization envelopes generated from Aarhus University Hospital Pharmacy, Denmark. The randomization numbers were ordered in blocks of 10 (five numbers assigned to tinzaparin treatment, five numbers assigned to no treatment). Participants assigned to tinzaparin treatment self-injected the trial drug subcutaneously in the periumbilical area. We blinded the outcome adjudicators to the allocation but not the participants and the treating doctors. For safety reasons, we measured plasma creatinine and platelet count at inclusion and we observed the participants for potential allergic reactions the first 30 min after the first drug injection. We repeated the platelet count after 1–2 weeks to ensure 39
Thrombosis Research 170 (2018) 38–44
A.T. Hansen et al.
Fig. 1. CONSORT flow diagram for the trial including singleton pregnancies complicated by foetal growth retardation.
statements 2010 [24,25].
study assignment. The extent of placental infarctions placental weight and presence of placental hypoplasia and of histological signs of uteroplacental hypo-perfusion were registered.
2.8. Statistical analysis 2.5. Biochemical testing We used an intention to treat analysis approach, including all randomized women in the statistical analyses [26]. We computed absolute mean differences for the primary outcome and secondary outcomes with 95% confidence intervals [95% CI]. For continuous variables, we used an unpaired Student's t-test, and for binominal variables, we used a chi-square test. Our study hypothesis was an increased birth weight in FGR pregnancies receiving tinzaparin compared to untreated pregnancies. The relative birth weight was expressed as the percentage of the expected birth weight as calculated from the gestational age at delivery and gender of the child [16]. We estimated the foetal growth rate for each patient by a linear regression analysis, using the ultrasound estimates of foetal weight and the birth weight. Next, the estimated growth rates in the two treatment groups were compared using linear regression. The flow measurements in the uterine, umbilical, and cerebral artery were analysed by mixed-model linear regression to allow for repeated measurements, adjusted for the gestational age at the time of measurement. All statistical analyses were performed using Stata 13.1 (StataCorp LP, Texas, USA). p-values < 0.05 were considered statistically significant. We performed an à priori sample size calculation. The clinically important difference in birth weight was considered 250 g [27]. We aimed for a 90% power with a significance level 0.05, and a standard deviation in mean births weight was set to 500 g leading to a sample size of 102 women in each group. During trial conduct, however, we realized that the distribution of mean birth weights in low birth weight infant is narrower than assumed in the initial sample size calculation [20]. Thus, the estimated number of women needed was 42 women in each group.
All biochemical analyses were performed at Department of Clinical Biochemistry, Aarhus University Hospital, Denmark, which is accredited as per international standards for laboratories (DS/EN ISO/IEC 15189). 2.6. Thrombophilia testing All participants underwent thorough thrombophilia testing. We genotyped for factor V Leiden variant (ARG506GLN) and the Prothrombin variant (20210G-A) and determined the activity of Protein C (Clot based functional assay) and Antithrombin (functional) (using STAR-R Evolution® (Stago Diagnostica, Asnières, France from start of trial to April 2013, and from April 2013 by CS 2100i (Sysmex, Kobe, Japan)). Antiphospholipid antibodies were determined three times during pregnancy. Lupus anticoagulant was measured by dilute Russell Viper Venom Time (dRVVT) (ACL TOP 700, Instrumentation Laboratory, Bedford, Massachusetts, USA and Siemens/Dade Behring, Marburg, Germany). A simplified dRVVT screened for lupus anticoagulant, and phospholipid rich dRVVT confirmed the presence of lupus anticoagulant in plasma using a ratio with a decision limit of ≥1.4. Anti- beta-2-glycoprotein 1 antibodies were until November 2012 measured by Asserchrom APA screen® kit (Stago Diagnostica, Asnières, France) and from December 2012–16 measured by Phadia 250 APS (Thermo Scientific, Uppsala, Sweden). 2.7. Biochemical evaluation of adherence to trial drug Consecutive measurements by anti-Xa-assay to determine 4-hour peak levels of anti-Xa were used as proxy for adherence to the trial drug. This randomized trial was reported according to the CONSORT 40
Thrombosis Research 170 (2018) 38–44
A.T. Hansen et al.
Table 2 Baseline characteristics of study participants (N = 53) at randomization to tinzaparin or no treatment for treatment of foetal growth retardation (FGR). Tinzaparin (N = 27)
No treatment (N = 26)
p-Value
Maternal age at inclusion, mean years, (SD) Pre-pregnancy body mass index, mean kg/m2, (SD) Gestational age at inclusion, mean days, (SD) Severity of FGR, mean deviation in % (SD)1 Active smoking, n (%) Gestational hypertension, n (%)2 Hyperemesis gravidarum, n (%) Conception by assisted reproduction, n (%)
33 22 202 −28 4 0 2 6
(6) (3) (15) (7) (15) (0) (7) (22)
32 22 193 −28 3 1 2 7
(6) (4) (26) (10) (12) (4) (8) (27)
0.08 0.74 0.12 0.99 0.73 0.30 0.97 0.69
Biochemical safety parameters Plasma creatinine, mean (SD), reference interval 45–90 μmol/L Platelet count, mean (SD), reference interval 165–400 × 109/L
50 227
(9) (49)
48 230
(9) (53)
0.46 0.41
Thrombophilia testing Heterozygote Factor V Leiden, n (%) Homozygote Factor V Leiden, n (%) Heterozygote Prothrombin variant, n (%) Homozygote Prothrombin variant, n (%) Lupus anticoagulant, n (%), at inclusion Lupus anticoagulant, n (%), at follow-up visit > 6 weeks Beta-2-glycoprotein-1-antibodies (IgG), n (%) Beta-2-glycoprotein-1-antibodies (IgG), n (%) > 6 weeks Protein C, mean × 103 IU/L (SD) Antithrombin, mean × 103 IU/L (SD)
4 0 0 0 7 7 1 0 1.10 1.05
(15) (0) (0) (0) (26) (26) (4) (0) (0.17) (0.11)
4 0 0 0 5 6 1 2 1.14 1.02
(15) (0) (0) (0) (19) (23) (4) (8) (0.24) (0.12)
0.95 – – – 0.40 0.22 0.37 0.41 0.52 0.38
Anamnestic obstetric history Previously pregnant, n (%) Foetal growth retardation, n (%) Still birth (after gestational week 24), n (%) Spontaneous abortion (before gestational week 24), n (%) Gestational hypertension, n (%)3 Preeclampsia, n (%)4
16 3 1 4 3 3
(59) (11) (4) (15) (11) (11)
20 0 2 4 1 1
(77) (0) (8) (15) (4) (4)
0.17 0.05 0.69 0.72 0.23 0.23
All percentages provided are per randomization group, unless otherwise stated. p-Values: For continuous variables, an unpaired student's t-test, was used and for binominal variables, chi-square test was used. SD: Standard deviation. Information on previous pregnancies was obtained during the systematic interview upon inclusion. 1 Estimated foetal weight measured using a standardized approach by obstetricians specialized in foetal medicine, based on foetal growth curves [16]. Stated as percentage deviation from expected at gestational age. 2 Functional measurements. Reference interval for Protein C: 0.74–1.50 × 103 IU/L. Reference interval for Antitrombin: 0.8–1.2 × 103 IU/L. 3 Gestational hypertension: hypertension after gestational weeks 20 (systolic/diastolic blood pressure above 140/90 mm Hg). 4 Preeclampsia: hypertension and proteinuria after gestational weeks 20.
3.3. Doppler flow-assessment in umbilical, uterine, and cerebral artery during follow-up
3. Results 3.1. Baseline data
Overall, we found no significant differences between the two randomized groups expressed for the Doppler assessment of flow in the umbilical, uterine, and cerebral arteries during follow-up. The umbilical artery pulsatility index declined during follow up in both treatment groups; at gestational age 30 weeks, it was 1.22 (95% CI: 1.16–1.27). In a mixed-model linear regression, allowing for repeated measurements and adjusting for gestational age, the pulsatility index was slightly lower in the tinzaparin treated group (difference: −0.09; 95% CI: −0.27, 0.08; p = 0.31). Similarly, there was no significant difference between treatment groups of the mid cerebral artery pulsatility index.
From November 2011 to August 2016, 53 women consented to participate. Of those, 27 were randomized to tinzaparin treatment and 26 were randomized to no treatment (Fig. 1). The baseline distribution at randomization of maternal age, body mass index, tobacco use, thrombophilia, and severity of FGR was balanced across the two groups (Table 2).
3.2. Birth weights and foetal growth rates The mean birth weight was 2229 g in the tinzaparin group compared to 1968 g in the no treatment group. However, the birth weight relative to the expected from gestational age and gender was only 2.5 percentage points higher in the tinzaparin group compared to the no treatment group [−5.1 to 10.0] (p = 0.51). The foetal growth rate during follow-up was 124 g/week in the tinzaparin group and 119 g/ week in the no treatment group, a difference of 5 g/week [−19 to 29] (p = 0.67) (Fig. 2). FGR infants in the tinzaparin group were born at a mean gestational age of 261 days compared with 246 days in the no treatment group, an absolute difference of 15 days (p = 0.06).
3.4. Other outcomes Two perinatal deaths occurred in the no treatment group whereas there were no perinatal deaths in the tinzaparin group. The cause of death for the two foetuses was 1) intra-uterine strangulation from the umbilical cord with secondary asphyxia in gestational week 34, and 2) intra-uterine death from extreme FGR; birth weight 325 g at gestational age 25 + 4, reverse flow in the umbilical artery and impaired flow in ductus venosus. The extent of placental infarctions was 2.4% [95% CI 0.4–4.4] in the tinzaparin group and 7.2% [95% CI 3.7–10.6] in the no treatment group (p = 0.03). There was no difference in the placental weights 41
Thrombosis Research 170 (2018) 38–44
A.T. Hansen et al.
Table 3 Estimates of primary and secondary outcomes for all patients randomized in the trial (N = 53). Primary outcomes
Birthweight, grams Birthweight, percent of expected Average foetal growth rate (grams/week)
Tinzaparin (N = 27)
No tinzaparin (N = 26)
Difference
Mean
(95% CI)
Mean
(95% CI)
Mean
(95% CI)
p-Value
2222 71.4% 124
(1961 to 2483) (66.6 to 76.2) (110 to 138)
1968 68.9% 119
(1608 to 2328) (62.8 to 75.1) (99 to 139)
254 2.5% 5
(−179 to 688) (−5.1 to 10) (−19 to 29)
0.24 0.51 0.67
Secondary outcomes
Foetal outcomes Gestational age at delivery, days Days from randomization until delivery Perinatal death1 Apgar score ≤ 7 after 5 min Admission to NICU Stay at NCIU, days (N = 29) Respiratory distress Intraventricular haemorrhage Maternal outcomes Gestational hypertension, preeclampsia or HELLP Placental abruption Maternal bleeding Antepartum bleeding requiring admission Postpartum bleeding5 Placental outcomes Patients with valid information, n Extent of placental infarction Placental weight, grams Placental hypoplasia Retro-placentar bleeding Anti-Xa measures Anti-Xa KIU/L at follow-up visits
Tinzaparin (N = 27)
No tinzaparin (N = 26)
Difference
Mean (95% CI) or n (%)
Mean (95% CI) or n (%)
Mean
(95% CI)
p-Value
261 60 0 0 14 9 6 1
(253 to 268) (53 to 66) (0%) (0%) (52%) (4 to 14) (22%) (4%)
246 54 2 1 15 13 8 0
(233 to 260) (39 to 68) (8%) (4%) (58%) (6 to 20) (33%) (0%)
15 6 −8% −4% −6% −4 −11% 4%
(−0.4 to 29) (−9 to 21) (−18 to 3) (−12 to 4) (−33 to 21) (−12 to 5) (−36 to 13) (−4 to 11)
0.06 0.43 0.39 0.28 0.67 0.29 0.38 0.34
9 1
(33%) (4%)
6 0
(23%) (0%)
10% 4%
(−14 to 34) (−4 to 12)
0.41 0.30
0 1
(0%) (4%)
1 1
(4%) (4%)
−4% 0%
(−11 to 4) (−10 to 10)
0.30 0.98
13 2.4% 304 11 0 Mean 0.26
(0.4 to 4.4) (273 to 335) (85%) (0%) (SD) (0.09)
19 7.2% 315 13 3 Mean 0.11
(3.7 to 10.6) (259 to 370) (69%) (16%) (SD) (0.02)
4.8% −10 16% −16%
(0.5 to 9.1) (−89 to 68) (−12 to 45) (−32 to 1)
0.03 0.79 0.30 0.13
0.15
(0.11 to 0.20)
< 0.001
Estimated fetal weight (grams)
All percentages provided are per randomization group, unless otherwise stated. Standard deviation (SD). NICU: neonatal intensive care unit. 1 Still birth or death within the first week after delivery. 2 Gestational hypertension: hypertension after gestational weeks 20 (systolic/diastolic blood pressure above 140/90 mm Hg). 3 Preeclampsia: hypertension and proteinuria after gestational weeks 20. 4 HELLP; Haemolysis, elevated liver enzymes, and low platelets, an aggravated clinical state of preeclampsia. 5 Blood loss > 500 mL (vaginal delivery) or > 1 L (caesarean section).
six weeks after the first positive value were referred to repeated measures of lupus anticoagulant and beta-2-glycoprotein–1 antibodies scheduled three months after delivery. We do not have access to these follow-up data. No foetal or maternal bleeding episodes or heparin induced thrombocytopenia occurred. The mean platelet count at first follow-up visit was 218 in the tinzaparin group and 219 in the no tinzaparin group. One woman in the tinzaparin group developed a postpartum bleeding, which was explained by HELLP and therefore unlikely to be related to the trial drug. Three women (11%) assigned to tinzaparin treatment cancelled the treatment within four weeks after randomization; one because she feared the self-injections, and the other two for personal causes. No cross overs occurred in the control group. We could confirm adherence to the trial drug by anti-Xa levels for women in the tinzaparin group being within the defined anti-Xa target levels (0.20–0.60 KIU/L). The anti-Xa assay used has a validated measurement range going down to 0.10 KIU/L. Because of a rather imprecise anti-Xa determination in the lower anti-Xa levels ( ± 23%) patients free from low molecular weight heparin will have an anti-Xa level around 0.10 KIU/L, which was the case in the control group.
4000
3000
2000
1000
Expected No treatment LMWH treatment
0 25
30
35
40
Gestational age at examination (weeks) Fig. 2. Development in foetal growth, separately for the tinzaparin group and no treatment group, plotted against the expected foetal weights according to gestational age [16].
across the two groups. In the tinzaparin group, 33% of mothers developed hypertensive disease (gestational hypertension, preeclampsia, or HELLP (haemolysis, elevated liver enzymes, and low platelets)) during participation, compared with 23% in the no treatment group (p = 0.41). These numbers are not surprising, since FGR and hypertensive disease co-exist in many cases. More women than expected had lupus anticoagulant (Table 2). Women with positive lupus anticoagulant at the second determination
4. Discussion 4.1. Main findings In this open-labelled randomized controlled trial of ongoing FGR pregnancies, we found no evidence of a tinzaparin effect on the foetal 42
Thrombosis Research 170 (2018) 38–44
A.T. Hansen et al.
bias of limited influence, as the timing of delivery was entirely a conference decision based on national guidelines, including objective measures of foetal growth, the biophysical profile, and the maternal status. The higher gestational age at delivery in the tinzaparin group was an important finding, since gestational age at delivery is the most important prognostic factor for neonatal and 2-year infant outcomes in FGR infants [32]. Existing studies have focused on prophylactic treatment with low molecular weight heparin in pregnancies at risk of FGR [8,9,10,11], but no previous trials have investigated low molecular weight heparin for the treatment of FGR already diagnosed in ongoing pregnancy. The results of the present trial challenge the use of low molecular weight heparin for FGR as we found no evidence of an improved placental function, foetal growth rate, or any lowering effect on the uterine artery resistance as previously indicated [5,7,12,13,14,15].
growth rate or the birth weight after adjustment for gestational age. However, due to the small study size, we cannot exclude a minor beneficial or harmful effect. The tendency for higher birth weights in the tinzaparin group disappeared when we adjusted for gestational age at delivery. For both groups, the foetal growth rates found were much lower than the normal foetal growth rates of 200 g/week around 200 to 250 gestational days (Marsal K, Acta Paediatrica 1996). The women that received tinzaparin delivered at a later gestational age. For women in both groups, the extent of placental infarction was considered physiological for the gestational age and considered to have had no influence of foetal growth. We found no significant tinzaparin effect on the utero-placental perfusion during follow-up. We found that a considerable proportion of women in both groups had one or two positive tests for antiphospholipid antibodies (lupus anticoagulant and/or beta-2-glycoprotein antibodies). Antiphospholipid antibodies may play a role for the development of FGR and other placenta-mediated complications as part of obstetric antiphospholipid syndrome, though their precise prevalence in FGR is not known [28]. The prevalence of antiphospholipid antibodies is, however, higher in preeclampsia pregnancies (14%) as compared with healthy pregnancies (7%) [29]. Women tested positive for antiphospholipid antibodies during the study were not excluded from the trial for receiving prophylactic low molecular weight heparin treatment, since persistent antiphospholipid antibodies should be measured twice with at least 12 weeks distance [30]. We re-tested the patients after six weeks, and women with persistent antiphospholipid antibodies were referred to a confirmatory retesting three months after delivery.
Ethics statement The Ethics Committee of Central Regional Denmark (M-2011-0042), the Danish Data Protection Agency and The Danish Health Authority approved the trial. Funding The trial was funded by The Danish Council for Independent Research (Grant no: 0602-02173B FSS), Central Region of Denmark, and Department of Clinical Medicine, Health, Aarhus University, Denmark. Leo Pharma kindly donated the trial drug, tinzaparin (Innohep®). The funders had no role in the design or conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, critical review, and final approval of the paper.
4.2. Strength and limitations There are several strengths to highlight for the present trial. The trial was conducted at three obstetric centres, all using the same strict enrolment criteria and accurate determination of gestational lengths by first trimester crown rump length measurements [18]. We achieved a balanced distribution of all measured confounders at baseline, and evaluated treatment effect by robust endpoints. The consecutive measures of foetal growth and flow in the uterine, umbilical, and cerebral arteries enabled us to determine whether tinzaparin had an effect on foetal growth, and placental function. The finding of only sparse placental infarctions in both randomization groups is intriguing, since placental infarctions and hypoxia have been considered important contributing etiological causes for FGR [3,4]. Furthermore, tinzaparin did not show any clinically significant effect on the degree of placental infarcts. This is in line with previous findings from a study on women at high risk of adverse pregnancy outcomes randomized to unfractionated heparin or no heparin [31]. This study could not demonstrate any heparin effect on regression or progression of placental mal-perfusion [31]. It was a limitation that there were three non-compliers in the tinzaparin group, for which reason the intention to treat approach may have underestimated any effects of tinzaparin [25]. The present trial was designed to evaluate birth weight independently as primary outcome, however for clinical purposes we furthermore evaluated birth weights adjusted for gestational age as well as foetal growth rates for the two groups separately. The study was challenged by very strict exclusion criteria and a low acceptance rate for inclusion in the trial. As we were unable to include the expected number of participants, we did not achieve as strong statistical power as intended. Thus, we cannot rule out the risk of overlooking a treatment effect. We ended enrolment of study participants due to end of funding. The open-labelled design without a placebo arm constituted the main limitation of the study: it was not practically possible for us to include a placebo group of women receiving only placebo medicine. This potential bias was especially important concerning the timing of deliveries; the obstetricians might be more liable to postpone delivery of women treated with tinzaparin. Nevertheless, we consider the risk of
Conflict of interests Anette Tarp Hansen, Puk Sandager, Olav Bjørn Petersen, Mette Ramsing, Svend Juul, Niels Uldbjerg, and Jannie Dalby Salvig reported no conflict of interest. Anne Mette Hvas has received speaker's fees from CSL Behring, Leo Pharma, Bayer, Astellas, Boehringer-Ingelheim, and Bristol-Myers Squibb, and unrestricted research support from Octapharma, CSL Behring and Leo Pharma. Authors' contributions Anette Tarp Hansen contributed substantially to the study's concept and design; inclusion and follow-up of participants, data analysis and interpretation of data; drafting of the paper; critical review and revision of the intellectual content; and final approval of the version to be published. Anne-Mette Hvas was overall responsible for the planning and conduct of the trial, and contributed substantially to study's concept and design; interpretation of data; critical review and revision of the intellectual content; and final approval of the version to be published. Svend Juul contributed substantially to data analysis and interpretation of data; critical review and revision of the intellectual content; and final approval of the version to be published. Puk Sandager, Jannie Dalby Salvig, Olav Bjørn Petersen, and Niels Uldbjerg contributed substantially to the study's concept and design; inclusion and follow-up of participants, interpretation of data; critical review and revision of the intellectual content; and final approval of the version to be published. Mette Ramsing performed the post-natal placenta pathological examinations and contributed substantially to interpretation of data; critical review and revision of the intellectual content; and final approval of the version to be published. Acknowledgements We sincerely thank all the women participating in our study. We 43
Thrombosis Research 170 (2018) 38–44
A.T. Hansen et al.
thank Vivi Bo Mogensen and Mai Stenulm Veirup for their huge laboratory assistance. We thank the following medical doctors at the three obstetric departments for all their efforts in enrolling study participants: Hanne Søndergaard Jensen, Inger Stornes, and Marianne Louise Vang Østergård at Departments of Obstetrics and Gynaecology at Randers and Herning Regional Hospitals, Denmark.
Ultrasound Med. Biol. 32 (9) (2006) 1431–1435. [13] S. Campbell, J. Diaz-Recasens, D.R. Griffin, et al., New Doppler technique for assessing uteroplacental blood flow, Lancet 1 (8326 Pt 1) (1983) 675–677. [14] O. Gomez, F. Figueras, J.M. Martinez, et al., Sequential changes in uterine artery blood flow pattern between the first and second trimesters of gestation in relation to pregnancy outcome, Ultrasound Obstet. Gynecol. 28 (6) (2006) 802–808. [15] O. Gomez, F. Figueras, S. Fernandez, et al., Reference ranges for uterine artery mean pulsatility index at 11–41 weeks of gestation, Ultrasound Obstet. Gynecol. 32 (2) (2008) 128–132. [16] K. Marsal, P.H. Persson, T. Larsen, H. Lilja, A. Selbing, B. Sultan, Intrauterine growth curves based on ultrasonically estimated foetal weights, Acta Paediatr. 85 (7) (1996) 843–848. [17] Thrombosis and embolism during pregnancy and the puerperium, reducing the risk (green-top guideline no. 37a). https://www.rcog.org.uk/en/guidelines-researchservices/guidelines/gtg37a/. Updated 2015. [18] The fetal medicine foundation. https://Fetalmedicine.org/training-n-certification/ certificates-of-competence. Assessed on the 13th of March 2017. [19] M. Parra-Cordero, C. Lees, H. Missfelder-Lobos, P. Seed, C. Harris, Fetal arterial and venous Doppler pulsatility index and time averaged velocity ranges, Prenat. Diagn. 27 (13) (2007) 1251–1257. [20] C. Lee, G. Lim, W.S. Kim, H.S. Han, Clinical characteristics and outcome of incidental atrial septal openings in very low birth weight infants, Neonatology 105 (2) (2014) 85–90. [22] F.P. Hadlock, R.L. Deter, R.B. Harrist, S.K. Park, Estimating fetal age: computerassisted analysis of multiple fetal growth parameters, Radiology 152 (2) (1984) 497–501. [23] T.Y. Khong, E.E. Mooney, I. Ariel, et al., Sampling and definitions of placental lesions: Amsterdam placental workshop group consensus statement, Arch. Pathol. Lab. Med. 140 (7) (2016) 698–713. [24] S. Hopewell, M. Clarke, D. Moher, et al., CONSORT for reporting randomised trials in journal and conference abstracts, Lancet 371 (9609) (2008) 281–283. [25] http://www.consort-statement.org/checklists/view/32-consort. Assessed 14 January 2017. [26] D. Moher, K.F. Schulz, D.G. Altman, The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomised trials, Lancet 357 (9263) (2001) 1191–1194. [27] L.C. Poon, N. Volpe, B. Muto, A. Syngelaki, K.H. Nicolaides, Birthweight with gestation and maternal characteristics in live births and stillbirths, Fetal Diagn. Ther. 32 (3) (2012) 156–165. [28] D. Garcia, D. Erkan, Diagnosis and management of the antiphospholipid syndrome, N. Engl. J. Med. 378 (21) (2018) 2010–2021. [29] R. Ferrer-Oliveras, E. Llurba, L. Cabero-Roura, J. Alijotas-Reig, Prevalence and clinical usefulness of antiphospholipid and anticofactor antibodies in different Spanish preeclampsia subsets, Lupus 21 (3) (2012) 257–263. [30] S. Miyakis, M.D. Lockshin, T. Atsumi, et al., International consensus statement on an update of the classification criteria for definite antiphospholipid syndrome (APS), J. Thromb. Haemost. 4 (2) (2006) 295–306. [31] R. D'Souza, S. Keating, M. Walker, S. Drewlo, J. Kingdom, Unfractionated heparin and placental pathology in high-risk pregnancies: secondary analysis of a pilot randomized controlled trial, Placenta 35 (10) (2014) 816–823. [32] T. Stampalija, B. Arabin, H. Wolf, C.M. Bilardo, C. Lees, TRUFFLE investigators. Is middle cerebral artery Doppler related to neonatal and 2-year infant outcome in early fetal growth restriction? Am. J. Obstet. Gynecol. 216 (2017) 521.
Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.thromres.2018.08.006. References [1] Stillbirth Collaborative Research Network Writing Group, Causes of death among stillbirths, JAMA 306 (22) (2011) 2459–2468. [2] V. Flenady, L. Koopmans, P. Middleton, et al., Major risk factors for stillbirth in high-income countries: a systematic review and meta-analysis, Lancet 377 (9774) (2011) 1331–1340. [3] D.J. Roberts, M.D. Post, The placenta in pre-eclampsia and intrauterine growth restriction, J. Clin. Pathol. 61 (12) (2008) 1254–1260. [4] M.G. Walker, B. Fitzgerald, S. Keating, J.G. Ray, R. Windrim, J.C. Kingdom, Sexspecific basis of severe placental dysfunction leading to extreme preterm delivery, Placenta 33 (7) (2012) 568–571. [5] I.A. Greer, C. Nelson-Piercy, Low-molecular-weight heparins for thromboprophylaxis and treatment of venous thromboembolism in pregnancy: a systematic review of safety and efficacy, Blood 106 (2) (2005) 401–407. [6] B.P. McDonnell, K. Glennon, A. McTiernan, et al., Adjustment of therapeutic LMWH to achieve specific target anti-FXa activity does not affect outcomes in pregnant patients with venous thromboembolism, J. Thromb. Thrombolysis 43 (2017) 105–111. [7] C. Nelson-Piercy, R. Powrie, J.Y. Borg, et al., Tinzaparin use in pregnancy: an international, retrospective study of the safety and efficacy profile, Eur. J. Obstet. Gynecol. Reprod. Biol. 159 (2) (2011) 293–299. [8] I. Martinelli, P. Ruggenenti, I. Cetin, et al., Heparin in pregnant women with previous placenta-mediated pregnancy complications: a prospective, randomized, multicenter, controlled clinical trial, Blood 119 (14) (2012) 3269–3275. [9] M.A. Rodger, W.M. Hague, J. Kingdom, et al., Antepartum dalteparin versus no antepartum dalteparin for the prevention of pregnancy complications in pregnant women with thrombophilia (TIPPS): a multinational open-label randomised trial, Lancet 384 (9955) (2014) 1673–1683. [10] E. Rey, P. Garneau, M. David, et al., Dalteparin for the prevention of recurrence of placental-mediated complications of pregnancy in women without thrombophilia: a pilot randomized controlled trial, J. Thromb. Haemost. 7 (1) (2009) 58–64. [11] M.A. Rodger, J.C. Gris, J.I. de Vries, et al., Low-molecular-weight heparin and recurrent placenta-mediated pregnancy complications: a meta-analysis of individual patient data from randomised controlled trials, Lancet (2016). [12] M. Torricelli, F.M. Reis, P. Florio, et al., Low-molecular-weight heparin improves the performance of uterine artery Doppler velocimetry to predict preeclampsia and small-for-gestational age infant in women with gestational hypertension,
44
Contents lists available at ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Full Length Article
Tinzaparin for the treatment of foetal growth retardation: An open-labelled randomized clinical trial☆
T
⁎
Anette Tarp Hansena,b,c, , Puk Sandagerb,d, Mette Ramsinge, Olav B. Petersenb,d, Jannie D. Salvigd, Svend Juulf, Niels Uldbjergb,d, Anne-Mette Hvasa,b a
Aarhus University Hospital, Centre for Haemophilia and Thrombosis, Department of Clinical Biochemistry, Denmark Aarhus University, Department of Clinical Medicine, Denmark c Aalborg University Hospital, Department of Clinical Biochemistry, Denmark d Aarhus University Hospital, Department of Obstetrics and Gynaecology, Denmark e Randers Regional Hospital, Department of Pathology, Denmark f Section for Epidemiology, Department of Public Health, Aarhus University, Denmark b
A R T I C LE I N FO
A B S T R A C T
Keywords: Birth weight Foetal growth retardation Heparin, low molecular weight Placenta, thrombosis Randomized controlled trial
Objective: Foetal growth retardation (FGR) is a leading cause of perinatal death and long-term harms at survivors. Placental infarction plays a role in FGR, yet, no trials have evaluated whether low molecular weight heparins increase birth weight in ongoing FGR pregnancies. Methods: An open-labelled randomized trial in Denmark during 2011–2016, including singleton pregnant women with FGR (estimated foetal weight < 2.3 percentile) diagnosed before gestational weeks 32. Participants were randomly assigned using sealed, blinded envelopes 1:1 to tinzaparin (4500 IU daily until 37 gestational weeks) or no tinzaparin. The primary outcomes were the observed birthweight relative to the expected for gestational age and gender, and foetal growth rates in the two trial groups evaluated by an intention to treat analysis. Results: We enrolled 53 women. The mean gestational age was 261 days in the tinzaparin group and 246 days in the no treatment group. The mean birth weight was 2229 g in the tinzaparin group compared to 1968 g in the no treatment group. However, the birth weight relative to the expected from gestational age and gender was only 2.5 percentage points higher in the tinzaparin group [−5.1 to 10.0] (p = 0.51). The foetal growth rate during follow-up was 124 g/week in the tinzaparin group and 119 g/week in the no treatment group, a difference of 5 g/week [−19 to 29] (p = 0.67). Two perinatal deaths both occurred in the no treatment group. Conclusion: We found no evidence of a tinzaparin effect on the foetal growth rate or the birth weight after adjustment for gestational age.
1. Introduction Worldwide, foetal growth retardation (FGR) is a leading cause of perinatal death and long-term harms for the survivors [1,2]. Furthermore, also the consequences of prematurity challenge these new-borns, as premature delivery is yet the only treatment option. A cornerstone among the pathophysiologic pathways leading to FGR is placental vasculopathy with placental infarction and impaired blood flow in the utero-placental vessels [3,4]. Therefore, anticoagulant drugs may constitute a principle of treatment and among these, low molecular weight heparins are the drugs of choice during pregnancy because they do not
cross the placenta, have a favourable safety profile for the mother, and a predictable anticoagulant effect [5,6,7]. When used as a prophylactic intervention in pregnant women with a prior FGR pregnancy, low molecular weight heparins seems to increase birth weight in subsequent pregnancies [8,9,10]. Furthermore, low molecular weight heparins decrease the uterine artery resistance in hypertensive pregnancies [5,7,12], a mechanism from which also FGR pregnancies may benefit [13,14,15]. However, identification of women at risk of developing FGR is difficult. Yet, there is no evidence on the effect of low molecular weight heparin treatment for FGR diagnosed in ongoing pregnancies. To
☆
Trial registration. ClinicalTrials.gov (EudraCT no. 2011-000818-20). Funding. The Danish Council for Independent Research (0602-02173B FSS) and Central Region of Denmark. Leo Pharma donated the trial drug. ⁎ Corresponding author at: Aalborg University Hospital, Department of Clinical Biochemistry, Hobrovej 18-22, 9000 Aalborg, Denmark. E-mail address: [email protected] (A.T. Hansen). https://doi.org/10.1016/j.thromres.2018.08.006 Received 19 May 2018; Received in revised form 12 July 2018; Accepted 8 August 2018 Available online 10 August 2018 0049-3848/ © 2018 Elsevier Ltd. All rights reserved.
Thrombosis Research 170 (2018) 38–44
A.T. Hansen et al.
address this gap in knowledge, we conducted a randomized trial testing if the low molecular weight heparin, tinzaparin, improves birth weight in ongoing pregnancies complicated by foetal growth retardation. We hypothesized that low molecular weight heparin increases foetal growth in ongoing pregnancies complicated by FGR.
Table 1 Inclusion and exclusion criteria for the trial. Inclusion criteria 1: Singleton pregnant 2: Able to read and understand the written informed consent 3: Foetal growth retardation; estimated foetal weight at least ≤ −22% of expected weight according to gestational age (≈2.3 percentile) with or without Pathologically increased resistance in the uterine arteries; mean pulsatility index > 1.7 standard deviations after gestational weeks 24. In the initial protocol (before trial conduct was commenced), pathologically increased resistance in the uterine arteries from gestational week 19–20 was a separate inclusion criteria.
2. Methods 2.1. Setting and study design We conducted an open-labelled randomized controlled trial at three obstetric centres in Denmark (Aarhus University Hospital, Regional Hospital Randers, and Regional Hospital Herning). The trial was monitored by the Good Clinical Practice Unit at Aarhus University Hospital, Denmark according to “Guideline for good clinical practice for trials on pharmacological products”, World Health Organization, WHO Technical Report Series no 850, 1995, annex 3. The trial was registered in ClinicalTrials.gov (EudraCT no. 2011-000818-20).
Exclusion criteria 1: Age < 18 years 2: Pre-pregnancy maternal body weight > 90 kg 3: Unable to read and understand the written informed consent 4: Renal insufficiency (plasma creatinine > 150 μ/L) 5: Pre-existing hypertension (systolic/diastolic blood pressure > 140/90 mm Hg) 6: Diabetes type I and II 7: Active inflammatory bowel disease 8: Severe heart disease 9: Abuse of alcohol or any drugs 10: Known bleeding disorder (von Willebrand disease, thrombocytopenia, haemophilia) 11: Ongoing treatment with vitamin K antagonists 12: Known allergy to low-molecular weight heparins 13: Previous heparin induced thrombocytopenia (type II) 14: Serious bleeding within the last month 15: Pre-existing indication for anticoagulant prophylaxis with low-molecular weight heparin in current pregnancy 16: Foetal chromosome anomaly 17: Severe foetal malformations 18: Contraindication for tinzaparin 19: Gestational week > 32 weeks
2.2. Participants In the period from November 2011 to August 2016, we included singleton pregnancies diagnosed with FGR before 32 gestational weeks. FGR was defined as an estimated foetal weight below the 2.3 percentile (−2.0 standard deviations (SD); −22% of expected weight corrected for gestational age) with or without increased uterine artery resistance [16]. Increased uterine artery resistance was defined as a mean pulsatility index > 1.7 after gestational weeks 24 [14,15]. We made use of fixed prophylactic dosage of tinzaparin in women with a pre-pregnancy weight between 50 and 90 kg; 4500 international units tinzaparin daily [17]. Therefore, women with a pre-pregnancy weight above 90 kg were not eligible for the present trial [17]. All women attending the Danish antenatal care programme undergo a first trimester risk calculation including ultrasound scan and biochemical markers with predictive value for Down's syndrome. At this ultrasound scan, the crown-rump length is determined as well as a precise gestational age. All participants had this first trimester scan including crown rump length and nuchal translucency measurements as well as risk calculation according to Foetal Medicine Foundation standards, and estimated due date was in all cases calculated from the crown rump length measurement [18]. All study criteria for inclusion and exclusion of trial candidates are shown in Table 1.
Criteria for exclusion from the trial 1: Development of thrombocytopenia, platelet count < 80 × 109/L 2: Allergic reactions to tinzaparin 3: Bleeding requiring hospital admission 4: Non-adherence to study protocol or withdrawal of written informed consent 6: Indication for treatment with a low-molecular weight heparin in current pregnancy 7: Identification of foetal chromosome anomalies 8: Identification of severe congenital foetal heart disease
identification of heparin-induced thrombocytopenia. Furthermore, we advised the obstetricians to wait at least 12 h from the last tinzaparin injection to performing a caesarean section or spinal or epidural analgesic procedures. The participants attended follow-up visits every second week, where an obstetrician specialized in foetal medicine [18] conducted a transabdominal scan. This scan included an estimate of the foetal weight and consecutive Doppler assessments of the utero-placental flow if considered necessary by the obstetrician according to local foetal-medicine guidelines. The Doppler assessment was evaluated by pulsatility indices for the umbilical, uterine, and cerebral arteries [15,19]. Furthermore, a research technician drew a blood sample and obtained an interview according to a systematic case report form; evaluation of adherence to the trial drug, possible side effects, and exclusion criteria. In Denmark, the ultrasound-based monitoring of foetal weights is performed using Marsal's growth charts expressed as percentage deviations in estimated foetal weights according to a gender-specific and gestational age-specific reference curve [16,20]. Data from all scans were stored at the Danish Foetal Medicine Database from which we obtained the estimated foetal weight deviation according to gestational age, expressed as percentage deviation from expected for gestational age [16,22]. All scans were conducted using GE Voluson E8 or E10 ultrasound scanners using curve-linear abdominal probes (GE, Milwaukee, USA). After delivery, placental pathology examination was performed following standard procedures [23]. The pathologist was blinded to
2.3. Primary outcome The primary outcomes were the observed birthweight relative to the expected for gestational age and gender, and foetal growth rates in the two trial groups. Secondary maternal and foetal outcomes are presented in Table 3. 2.4. Trial drug and randomization procedure Upon inclusion, we randomized participants in a 1:1 ratio to either tinzaparin (Innohep®, Leo Pharma, subcutaneous self-injection of 4500 international units daily until completed 37 weeks of gestation) or no tinzaparin (standard procedure), using sealed and blinded randomization envelopes generated from Aarhus University Hospital Pharmacy, Denmark. The randomization numbers were ordered in blocks of 10 (five numbers assigned to tinzaparin treatment, five numbers assigned to no treatment). Participants assigned to tinzaparin treatment self-injected the trial drug subcutaneously in the periumbilical area. We blinded the outcome adjudicators to the allocation but not the participants and the treating doctors. For safety reasons, we measured plasma creatinine and platelet count at inclusion and we observed the participants for potential allergic reactions the first 30 min after the first drug injection. We repeated the platelet count after 1–2 weeks to ensure 39
Thrombosis Research 170 (2018) 38–44
A.T. Hansen et al.
Fig. 1. CONSORT flow diagram for the trial including singleton pregnancies complicated by foetal growth retardation.
statements 2010 [24,25].
study assignment. The extent of placental infarctions placental weight and presence of placental hypoplasia and of histological signs of uteroplacental hypo-perfusion were registered.
2.8. Statistical analysis 2.5. Biochemical testing We used an intention to treat analysis approach, including all randomized women in the statistical analyses [26]. We computed absolute mean differences for the primary outcome and secondary outcomes with 95% confidence intervals [95% CI]. For continuous variables, we used an unpaired Student's t-test, and for binominal variables, we used a chi-square test. Our study hypothesis was an increased birth weight in FGR pregnancies receiving tinzaparin compared to untreated pregnancies. The relative birth weight was expressed as the percentage of the expected birth weight as calculated from the gestational age at delivery and gender of the child [16]. We estimated the foetal growth rate for each patient by a linear regression analysis, using the ultrasound estimates of foetal weight and the birth weight. Next, the estimated growth rates in the two treatment groups were compared using linear regression. The flow measurements in the uterine, umbilical, and cerebral artery were analysed by mixed-model linear regression to allow for repeated measurements, adjusted for the gestational age at the time of measurement. All statistical analyses were performed using Stata 13.1 (StataCorp LP, Texas, USA). p-values < 0.05 were considered statistically significant. We performed an à priori sample size calculation. The clinically important difference in birth weight was considered 250 g [27]. We aimed for a 90% power with a significance level 0.05, and a standard deviation in mean births weight was set to 500 g leading to a sample size of 102 women in each group. During trial conduct, however, we realized that the distribution of mean birth weights in low birth weight infant is narrower than assumed in the initial sample size calculation [20]. Thus, the estimated number of women needed was 42 women in each group.
All biochemical analyses were performed at Department of Clinical Biochemistry, Aarhus University Hospital, Denmark, which is accredited as per international standards for laboratories (DS/EN ISO/IEC 15189). 2.6. Thrombophilia testing All participants underwent thorough thrombophilia testing. We genotyped for factor V Leiden variant (ARG506GLN) and the Prothrombin variant (20210G-A) and determined the activity of Protein C (Clot based functional assay) and Antithrombin (functional) (using STAR-R Evolution® (Stago Diagnostica, Asnières, France from start of trial to April 2013, and from April 2013 by CS 2100i (Sysmex, Kobe, Japan)). Antiphospholipid antibodies were determined three times during pregnancy. Lupus anticoagulant was measured by dilute Russell Viper Venom Time (dRVVT) (ACL TOP 700, Instrumentation Laboratory, Bedford, Massachusetts, USA and Siemens/Dade Behring, Marburg, Germany). A simplified dRVVT screened for lupus anticoagulant, and phospholipid rich dRVVT confirmed the presence of lupus anticoagulant in plasma using a ratio with a decision limit of ≥1.4. Anti- beta-2-glycoprotein 1 antibodies were until November 2012 measured by Asserchrom APA screen® kit (Stago Diagnostica, Asnières, France) and from December 2012–16 measured by Phadia 250 APS (Thermo Scientific, Uppsala, Sweden). 2.7. Biochemical evaluation of adherence to trial drug Consecutive measurements by anti-Xa-assay to determine 4-hour peak levels of anti-Xa were used as proxy for adherence to the trial drug. This randomized trial was reported according to the CONSORT 40
Thrombosis Research 170 (2018) 38–44
A.T. Hansen et al.
Table 2 Baseline characteristics of study participants (N = 53) at randomization to tinzaparin or no treatment for treatment of foetal growth retardation (FGR). Tinzaparin (N = 27)
No treatment (N = 26)
p-Value
Maternal age at inclusion, mean years, (SD) Pre-pregnancy body mass index, mean kg/m2, (SD) Gestational age at inclusion, mean days, (SD) Severity of FGR, mean deviation in % (SD)1 Active smoking, n (%) Gestational hypertension, n (%)2 Hyperemesis gravidarum, n (%) Conception by assisted reproduction, n (%)
33 22 202 −28 4 0 2 6
(6) (3) (15) (7) (15) (0) (7) (22)
32 22 193 −28 3 1 2 7
(6) (4) (26) (10) (12) (4) (8) (27)
0.08 0.74 0.12 0.99 0.73 0.30 0.97 0.69
Biochemical safety parameters Plasma creatinine, mean (SD), reference interval 45–90 μmol/L Platelet count, mean (SD), reference interval 165–400 × 109/L
50 227
(9) (49)
48 230
(9) (53)
0.46 0.41
Thrombophilia testing Heterozygote Factor V Leiden, n (%) Homozygote Factor V Leiden, n (%) Heterozygote Prothrombin variant, n (%) Homozygote Prothrombin variant, n (%) Lupus anticoagulant, n (%), at inclusion Lupus anticoagulant, n (%), at follow-up visit > 6 weeks Beta-2-glycoprotein-1-antibodies (IgG), n (%) Beta-2-glycoprotein-1-antibodies (IgG), n (%) > 6 weeks Protein C, mean × 103 IU/L (SD) Antithrombin, mean × 103 IU/L (SD)
4 0 0 0 7 7 1 0 1.10 1.05
(15) (0) (0) (0) (26) (26) (4) (0) (0.17) (0.11)
4 0 0 0 5 6 1 2 1.14 1.02
(15) (0) (0) (0) (19) (23) (4) (8) (0.24) (0.12)
0.95 – – – 0.40 0.22 0.37 0.41 0.52 0.38
Anamnestic obstetric history Previously pregnant, n (%) Foetal growth retardation, n (%) Still birth (after gestational week 24), n (%) Spontaneous abortion (before gestational week 24), n (%) Gestational hypertension, n (%)3 Preeclampsia, n (%)4
16 3 1 4 3 3
(59) (11) (4) (15) (11) (11)
20 0 2 4 1 1
(77) (0) (8) (15) (4) (4)
0.17 0.05 0.69 0.72 0.23 0.23
All percentages provided are per randomization group, unless otherwise stated. p-Values: For continuous variables, an unpaired student's t-test, was used and for binominal variables, chi-square test was used. SD: Standard deviation. Information on previous pregnancies was obtained during the systematic interview upon inclusion. 1 Estimated foetal weight measured using a standardized approach by obstetricians specialized in foetal medicine, based on foetal growth curves [16]. Stated as percentage deviation from expected at gestational age. 2 Functional measurements. Reference interval for Protein C: 0.74–1.50 × 103 IU/L. Reference interval for Antitrombin: 0.8–1.2 × 103 IU/L. 3 Gestational hypertension: hypertension after gestational weeks 20 (systolic/diastolic blood pressure above 140/90 mm Hg). 4 Preeclampsia: hypertension and proteinuria after gestational weeks 20.
3.3. Doppler flow-assessment in umbilical, uterine, and cerebral artery during follow-up
3. Results 3.1. Baseline data
Overall, we found no significant differences between the two randomized groups expressed for the Doppler assessment of flow in the umbilical, uterine, and cerebral arteries during follow-up. The umbilical artery pulsatility index declined during follow up in both treatment groups; at gestational age 30 weeks, it was 1.22 (95% CI: 1.16–1.27). In a mixed-model linear regression, allowing for repeated measurements and adjusting for gestational age, the pulsatility index was slightly lower in the tinzaparin treated group (difference: −0.09; 95% CI: −0.27, 0.08; p = 0.31). Similarly, there was no significant difference between treatment groups of the mid cerebral artery pulsatility index.
From November 2011 to August 2016, 53 women consented to participate. Of those, 27 were randomized to tinzaparin treatment and 26 were randomized to no treatment (Fig. 1). The baseline distribution at randomization of maternal age, body mass index, tobacco use, thrombophilia, and severity of FGR was balanced across the two groups (Table 2).
3.2. Birth weights and foetal growth rates The mean birth weight was 2229 g in the tinzaparin group compared to 1968 g in the no treatment group. However, the birth weight relative to the expected from gestational age and gender was only 2.5 percentage points higher in the tinzaparin group compared to the no treatment group [−5.1 to 10.0] (p = 0.51). The foetal growth rate during follow-up was 124 g/week in the tinzaparin group and 119 g/ week in the no treatment group, a difference of 5 g/week [−19 to 29] (p = 0.67) (Fig. 2). FGR infants in the tinzaparin group were born at a mean gestational age of 261 days compared with 246 days in the no treatment group, an absolute difference of 15 days (p = 0.06).
3.4. Other outcomes Two perinatal deaths occurred in the no treatment group whereas there were no perinatal deaths in the tinzaparin group. The cause of death for the two foetuses was 1) intra-uterine strangulation from the umbilical cord with secondary asphyxia in gestational week 34, and 2) intra-uterine death from extreme FGR; birth weight 325 g at gestational age 25 + 4, reverse flow in the umbilical artery and impaired flow in ductus venosus. The extent of placental infarctions was 2.4% [95% CI 0.4–4.4] in the tinzaparin group and 7.2% [95% CI 3.7–10.6] in the no treatment group (p = 0.03). There was no difference in the placental weights 41
Thrombosis Research 170 (2018) 38–44
A.T. Hansen et al.
Table 3 Estimates of primary and secondary outcomes for all patients randomized in the trial (N = 53). Primary outcomes
Birthweight, grams Birthweight, percent of expected Average foetal growth rate (grams/week)
Tinzaparin (N = 27)
No tinzaparin (N = 26)
Difference
Mean
(95% CI)
Mean
(95% CI)
Mean
(95% CI)
p-Value
2222 71.4% 124
(1961 to 2483) (66.6 to 76.2) (110 to 138)
1968 68.9% 119
(1608 to 2328) (62.8 to 75.1) (99 to 139)
254 2.5% 5
(−179 to 688) (−5.1 to 10) (−19 to 29)
0.24 0.51 0.67
Secondary outcomes
Foetal outcomes Gestational age at delivery, days Days from randomization until delivery Perinatal death1 Apgar score ≤ 7 after 5 min Admission to NICU Stay at NCIU, days (N = 29) Respiratory distress Intraventricular haemorrhage Maternal outcomes Gestational hypertension, preeclampsia or HELLP Placental abruption Maternal bleeding Antepartum bleeding requiring admission Postpartum bleeding5 Placental outcomes Patients with valid information, n Extent of placental infarction Placental weight, grams Placental hypoplasia Retro-placentar bleeding Anti-Xa measures Anti-Xa KIU/L at follow-up visits
Tinzaparin (N = 27)
No tinzaparin (N = 26)
Difference
Mean (95% CI) or n (%)
Mean (95% CI) or n (%)
Mean
(95% CI)
p-Value
261 60 0 0 14 9 6 1
(253 to 268) (53 to 66) (0%) (0%) (52%) (4 to 14) (22%) (4%)
246 54 2 1 15 13 8 0
(233 to 260) (39 to 68) (8%) (4%) (58%) (6 to 20) (33%) (0%)
15 6 −8% −4% −6% −4 −11% 4%
(−0.4 to 29) (−9 to 21) (−18 to 3) (−12 to 4) (−33 to 21) (−12 to 5) (−36 to 13) (−4 to 11)
0.06 0.43 0.39 0.28 0.67 0.29 0.38 0.34
9 1
(33%) (4%)
6 0
(23%) (0%)
10% 4%
(−14 to 34) (−4 to 12)
0.41 0.30
0 1
(0%) (4%)
1 1
(4%) (4%)
−4% 0%
(−11 to 4) (−10 to 10)
0.30 0.98
13 2.4% 304 11 0 Mean 0.26
(0.4 to 4.4) (273 to 335) (85%) (0%) (SD) (0.09)
19 7.2% 315 13 3 Mean 0.11
(3.7 to 10.6) (259 to 370) (69%) (16%) (SD) (0.02)
4.8% −10 16% −16%
(0.5 to 9.1) (−89 to 68) (−12 to 45) (−32 to 1)
0.03 0.79 0.30 0.13
0.15
(0.11 to 0.20)
< 0.001
Estimated fetal weight (grams)
All percentages provided are per randomization group, unless otherwise stated. Standard deviation (SD). NICU: neonatal intensive care unit. 1 Still birth or death within the first week after delivery. 2 Gestational hypertension: hypertension after gestational weeks 20 (systolic/diastolic blood pressure above 140/90 mm Hg). 3 Preeclampsia: hypertension and proteinuria after gestational weeks 20. 4 HELLP; Haemolysis, elevated liver enzymes, and low platelets, an aggravated clinical state of preeclampsia. 5 Blood loss > 500 mL (vaginal delivery) or > 1 L (caesarean section).
six weeks after the first positive value were referred to repeated measures of lupus anticoagulant and beta-2-glycoprotein–1 antibodies scheduled three months after delivery. We do not have access to these follow-up data. No foetal or maternal bleeding episodes or heparin induced thrombocytopenia occurred. The mean platelet count at first follow-up visit was 218 in the tinzaparin group and 219 in the no tinzaparin group. One woman in the tinzaparin group developed a postpartum bleeding, which was explained by HELLP and therefore unlikely to be related to the trial drug. Three women (11%) assigned to tinzaparin treatment cancelled the treatment within four weeks after randomization; one because she feared the self-injections, and the other two for personal causes. No cross overs occurred in the control group. We could confirm adherence to the trial drug by anti-Xa levels for women in the tinzaparin group being within the defined anti-Xa target levels (0.20–0.60 KIU/L). The anti-Xa assay used has a validated measurement range going down to 0.10 KIU/L. Because of a rather imprecise anti-Xa determination in the lower anti-Xa levels ( ± 23%) patients free from low molecular weight heparin will have an anti-Xa level around 0.10 KIU/L, which was the case in the control group.
4000
3000
2000
1000
Expected No treatment LMWH treatment
0 25
30
35
40
Gestational age at examination (weeks) Fig. 2. Development in foetal growth, separately for the tinzaparin group and no treatment group, plotted against the expected foetal weights according to gestational age [16].
across the two groups. In the tinzaparin group, 33% of mothers developed hypertensive disease (gestational hypertension, preeclampsia, or HELLP (haemolysis, elevated liver enzymes, and low platelets)) during participation, compared with 23% in the no treatment group (p = 0.41). These numbers are not surprising, since FGR and hypertensive disease co-exist in many cases. More women than expected had lupus anticoagulant (Table 2). Women with positive lupus anticoagulant at the second determination
4. Discussion 4.1. Main findings In this open-labelled randomized controlled trial of ongoing FGR pregnancies, we found no evidence of a tinzaparin effect on the foetal 42
Thrombosis Research 170 (2018) 38–44
A.T. Hansen et al.
bias of limited influence, as the timing of delivery was entirely a conference decision based on national guidelines, including objective measures of foetal growth, the biophysical profile, and the maternal status. The higher gestational age at delivery in the tinzaparin group was an important finding, since gestational age at delivery is the most important prognostic factor for neonatal and 2-year infant outcomes in FGR infants [32]. Existing studies have focused on prophylactic treatment with low molecular weight heparin in pregnancies at risk of FGR [8,9,10,11], but no previous trials have investigated low molecular weight heparin for the treatment of FGR already diagnosed in ongoing pregnancy. The results of the present trial challenge the use of low molecular weight heparin for FGR as we found no evidence of an improved placental function, foetal growth rate, or any lowering effect on the uterine artery resistance as previously indicated [5,7,12,13,14,15].
growth rate or the birth weight after adjustment for gestational age. However, due to the small study size, we cannot exclude a minor beneficial or harmful effect. The tendency for higher birth weights in the tinzaparin group disappeared when we adjusted for gestational age at delivery. For both groups, the foetal growth rates found were much lower than the normal foetal growth rates of 200 g/week around 200 to 250 gestational days (Marsal K, Acta Paediatrica 1996). The women that received tinzaparin delivered at a later gestational age. For women in both groups, the extent of placental infarction was considered physiological for the gestational age and considered to have had no influence of foetal growth. We found no significant tinzaparin effect on the utero-placental perfusion during follow-up. We found that a considerable proportion of women in both groups had one or two positive tests for antiphospholipid antibodies (lupus anticoagulant and/or beta-2-glycoprotein antibodies). Antiphospholipid antibodies may play a role for the development of FGR and other placenta-mediated complications as part of obstetric antiphospholipid syndrome, though their precise prevalence in FGR is not known [28]. The prevalence of antiphospholipid antibodies is, however, higher in preeclampsia pregnancies (14%) as compared with healthy pregnancies (7%) [29]. Women tested positive for antiphospholipid antibodies during the study were not excluded from the trial for receiving prophylactic low molecular weight heparin treatment, since persistent antiphospholipid antibodies should be measured twice with at least 12 weeks distance [30]. We re-tested the patients after six weeks, and women with persistent antiphospholipid antibodies were referred to a confirmatory retesting three months after delivery.
Ethics statement The Ethics Committee of Central Regional Denmark (M-2011-0042), the Danish Data Protection Agency and The Danish Health Authority approved the trial. Funding The trial was funded by The Danish Council for Independent Research (Grant no: 0602-02173B FSS), Central Region of Denmark, and Department of Clinical Medicine, Health, Aarhus University, Denmark. Leo Pharma kindly donated the trial drug, tinzaparin (Innohep®). The funders had no role in the design or conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, critical review, and final approval of the paper.
4.2. Strength and limitations There are several strengths to highlight for the present trial. The trial was conducted at three obstetric centres, all using the same strict enrolment criteria and accurate determination of gestational lengths by first trimester crown rump length measurements [18]. We achieved a balanced distribution of all measured confounders at baseline, and evaluated treatment effect by robust endpoints. The consecutive measures of foetal growth and flow in the uterine, umbilical, and cerebral arteries enabled us to determine whether tinzaparin had an effect on foetal growth, and placental function. The finding of only sparse placental infarctions in both randomization groups is intriguing, since placental infarctions and hypoxia have been considered important contributing etiological causes for FGR [3,4]. Furthermore, tinzaparin did not show any clinically significant effect on the degree of placental infarcts. This is in line with previous findings from a study on women at high risk of adverse pregnancy outcomes randomized to unfractionated heparin or no heparin [31]. This study could not demonstrate any heparin effect on regression or progression of placental mal-perfusion [31]. It was a limitation that there were three non-compliers in the tinzaparin group, for which reason the intention to treat approach may have underestimated any effects of tinzaparin [25]. The present trial was designed to evaluate birth weight independently as primary outcome, however for clinical purposes we furthermore evaluated birth weights adjusted for gestational age as well as foetal growth rates for the two groups separately. The study was challenged by very strict exclusion criteria and a low acceptance rate for inclusion in the trial. As we were unable to include the expected number of participants, we did not achieve as strong statistical power as intended. Thus, we cannot rule out the risk of overlooking a treatment effect. We ended enrolment of study participants due to end of funding. The open-labelled design without a placebo arm constituted the main limitation of the study: it was not practically possible for us to include a placebo group of women receiving only placebo medicine. This potential bias was especially important concerning the timing of deliveries; the obstetricians might be more liable to postpone delivery of women treated with tinzaparin. Nevertheless, we consider the risk of
Conflict of interests Anette Tarp Hansen, Puk Sandager, Olav Bjørn Petersen, Mette Ramsing, Svend Juul, Niels Uldbjerg, and Jannie Dalby Salvig reported no conflict of interest. Anne Mette Hvas has received speaker's fees from CSL Behring, Leo Pharma, Bayer, Astellas, Boehringer-Ingelheim, and Bristol-Myers Squibb, and unrestricted research support from Octapharma, CSL Behring and Leo Pharma. Authors' contributions Anette Tarp Hansen contributed substantially to the study's concept and design; inclusion and follow-up of participants, data analysis and interpretation of data; drafting of the paper; critical review and revision of the intellectual content; and final approval of the version to be published. Anne-Mette Hvas was overall responsible for the planning and conduct of the trial, and contributed substantially to study's concept and design; interpretation of data; critical review and revision of the intellectual content; and final approval of the version to be published. Svend Juul contributed substantially to data analysis and interpretation of data; critical review and revision of the intellectual content; and final approval of the version to be published. Puk Sandager, Jannie Dalby Salvig, Olav Bjørn Petersen, and Niels Uldbjerg contributed substantially to the study's concept and design; inclusion and follow-up of participants, interpretation of data; critical review and revision of the intellectual content; and final approval of the version to be published. Mette Ramsing performed the post-natal placenta pathological examinations and contributed substantially to interpretation of data; critical review and revision of the intellectual content; and final approval of the version to be published. Acknowledgements We sincerely thank all the women participating in our study. We 43
Thrombosis Research 170 (2018) 38–44
A.T. Hansen et al.
thank Vivi Bo Mogensen and Mai Stenulm Veirup for their huge laboratory assistance. We thank the following medical doctors at the three obstetric departments for all their efforts in enrolling study participants: Hanne Søndergaard Jensen, Inger Stornes, and Marianne Louise Vang Østergård at Departments of Obstetrics and Gynaecology at Randers and Herning Regional Hospitals, Denmark.
Ultrasound Med. Biol. 32 (9) (2006) 1431–1435. [13] S. Campbell, J. Diaz-Recasens, D.R. Griffin, et al., New Doppler technique for assessing uteroplacental blood flow, Lancet 1 (8326 Pt 1) (1983) 675–677. [14] O. Gomez, F. Figueras, J.M. Martinez, et al., Sequential changes in uterine artery blood flow pattern between the first and second trimesters of gestation in relation to pregnancy outcome, Ultrasound Obstet. Gynecol. 28 (6) (2006) 802–808. [15] O. Gomez, F. Figueras, S. Fernandez, et al., Reference ranges for uterine artery mean pulsatility index at 11–41 weeks of gestation, Ultrasound Obstet. Gynecol. 32 (2) (2008) 128–132. [16] K. Marsal, P.H. Persson, T. Larsen, H. Lilja, A. Selbing, B. Sultan, Intrauterine growth curves based on ultrasonically estimated foetal weights, Acta Paediatr. 85 (7) (1996) 843–848. [17] Thrombosis and embolism during pregnancy and the puerperium, reducing the risk (green-top guideline no. 37a). https://www.rcog.org.uk/en/guidelines-researchservices/guidelines/gtg37a/. Updated 2015. [18] The fetal medicine foundation. https://Fetalmedicine.org/training-n-certification/ certificates-of-competence. Assessed on the 13th of March 2017. [19] M. Parra-Cordero, C. Lees, H. Missfelder-Lobos, P. Seed, C. Harris, Fetal arterial and venous Doppler pulsatility index and time averaged velocity ranges, Prenat. Diagn. 27 (13) (2007) 1251–1257. [20] C. Lee, G. Lim, W.S. Kim, H.S. Han, Clinical characteristics and outcome of incidental atrial septal openings in very low birth weight infants, Neonatology 105 (2) (2014) 85–90. [22] F.P. Hadlock, R.L. Deter, R.B. Harrist, S.K. Park, Estimating fetal age: computerassisted analysis of multiple fetal growth parameters, Radiology 152 (2) (1984) 497–501. [23] T.Y. Khong, E.E. Mooney, I. Ariel, et al., Sampling and definitions of placental lesions: Amsterdam placental workshop group consensus statement, Arch. Pathol. Lab. Med. 140 (7) (2016) 698–713. [24] S. Hopewell, M. Clarke, D. Moher, et al., CONSORT for reporting randomised trials in journal and conference abstracts, Lancet 371 (9609) (2008) 281–283. [25] http://www.consort-statement.org/checklists/view/32-consort. Assessed 14 January 2017. [26] D. Moher, K.F. Schulz, D.G. Altman, The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomised trials, Lancet 357 (9263) (2001) 1191–1194. [27] L.C. Poon, N. Volpe, B. Muto, A. Syngelaki, K.H. Nicolaides, Birthweight with gestation and maternal characteristics in live births and stillbirths, Fetal Diagn. Ther. 32 (3) (2012) 156–165. [28] D. Garcia, D. Erkan, Diagnosis and management of the antiphospholipid syndrome, N. Engl. J. Med. 378 (21) (2018) 2010–2021. [29] R. Ferrer-Oliveras, E. Llurba, L. Cabero-Roura, J. Alijotas-Reig, Prevalence and clinical usefulness of antiphospholipid and anticofactor antibodies in different Spanish preeclampsia subsets, Lupus 21 (3) (2012) 257–263. [30] S. Miyakis, M.D. Lockshin, T. Atsumi, et al., International consensus statement on an update of the classification criteria for definite antiphospholipid syndrome (APS), J. Thromb. Haemost. 4 (2) (2006) 295–306. [31] R. D'Souza, S. Keating, M. Walker, S. Drewlo, J. Kingdom, Unfractionated heparin and placental pathology in high-risk pregnancies: secondary analysis of a pilot randomized controlled trial, Placenta 35 (10) (2014) 816–823. [32] T. Stampalija, B. Arabin, H. Wolf, C.M. Bilardo, C. Lees, TRUFFLE investigators. Is middle cerebral artery Doppler related to neonatal and 2-year infant outcome in early fetal growth restriction? Am. J. Obstet. Gynecol. 216 (2017) 521.
Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.thromres.2018.08.006. References [1] Stillbirth Collaborative Research Network Writing Group, Causes of death among stillbirths, JAMA 306 (22) (2011) 2459–2468. [2] V. Flenady, L. Koopmans, P. Middleton, et al., Major risk factors for stillbirth in high-income countries: a systematic review and meta-analysis, Lancet 377 (9774) (2011) 1331–1340. [3] D.J. Roberts, M.D. Post, The placenta in pre-eclampsia and intrauterine growth restriction, J. Clin. Pathol. 61 (12) (2008) 1254–1260. [4] M.G. Walker, B. Fitzgerald, S. Keating, J.G. Ray, R. Windrim, J.C. Kingdom, Sexspecific basis of severe placental dysfunction leading to extreme preterm delivery, Placenta 33 (7) (2012) 568–571. [5] I.A. Greer, C. Nelson-Piercy, Low-molecular-weight heparins for thromboprophylaxis and treatment of venous thromboembolism in pregnancy: a systematic review of safety and efficacy, Blood 106 (2) (2005) 401–407. [6] B.P. McDonnell, K. Glennon, A. McTiernan, et al., Adjustment of therapeutic LMWH to achieve specific target anti-FXa activity does not affect outcomes in pregnant patients with venous thromboembolism, J. Thromb. Thrombolysis 43 (2017) 105–111. [7] C. Nelson-Piercy, R. Powrie, J.Y. Borg, et al., Tinzaparin use in pregnancy: an international, retrospective study of the safety and efficacy profile, Eur. J. Obstet. Gynecol. Reprod. Biol. 159 (2) (2011) 293–299. [8] I. Martinelli, P. Ruggenenti, I. Cetin, et al., Heparin in pregnant women with previous placenta-mediated pregnancy complications: a prospective, randomized, multicenter, controlled clinical trial, Blood 119 (14) (2012) 3269–3275. [9] M.A. Rodger, W.M. Hague, J. Kingdom, et al., Antepartum dalteparin versus no antepartum dalteparin for the prevention of pregnancy complications in pregnant women with thrombophilia (TIPPS): a multinational open-label randomised trial, Lancet 384 (9955) (2014) 1673–1683. [10] E. Rey, P. Garneau, M. David, et al., Dalteparin for the prevention of recurrence of placental-mediated complications of pregnancy in women without thrombophilia: a pilot randomized controlled trial, J. Thromb. Haemost. 7 (1) (2009) 58–64. [11] M.A. Rodger, J.C. Gris, J.I. de Vries, et al., Low-molecular-weight heparin and recurrent placenta-mediated pregnancy complications: a meta-analysis of individual patient data from randomised controlled trials, Lancet (2016). [12] M. Torricelli, F.M. Reis, P. Florio, et al., Low-molecular-weight heparin improves the performance of uterine artery Doppler velocimetry to predict preeclampsia and small-for-gestational age infant in women with gestational hypertension,
44