Safety of Pregnancy After Cerebral Venous Thrombosis
Results of the ISCVT (International Study on Cerebral Vein and Dural Sinus Thrombosis)-2 PREGNANCY Study
Background and Purpose—Pregnancy is associated with increased risk of venous thrombotic events, including cerebral venous thrombosis. We aimed to study the complications and outcome of subsequent pregnancies in women with previous cerebral venous thrombosis.
Methods—Follow-up study of women with acute cerebral venous thrombosis at childbearing age included in a previously described cohort (International Study of Cerebral Vein and Dural Sinus Thrombosis). Patients were interviewed by local neurologists to assess rate of venous thrombotic events, pregnancy outcomes, and antithrombotic prophylaxis during subsequent pregnancies.
Results—A total of 119 women were included, with a median follow-up of 14 years. Eighty-two new pregnancies occurred in 47 women. In 83% (68 of 82), some form of antithrombotic prophylaxis was given during at least 1 trimester of pregnancy or puerperium. Venous thrombotic events occurred in 3 pregnancies, including 1 recurrent cerebral venous thrombosis. Two of the 3 women were on prophylactic low-molecular-weight heparin at the time of the event. Outcomes of pregnancies were 51 full-term newborns, 9 preterm births, 2 stillbirths, and 20 abortions (14 spontaneous).
Conclusions—In women with prior cerebral venous thrombosis, recurrent venous thrombotic events during subsequent pregnancies are infrequent.
- follow-up studies
- infant, newborn
- postpartum period
- secondary prevention
- sinus thrombosis
Pregnancy and puerperium are associated with an increased risk of venous thrombotic events (VTE). Eight to 58% of cerebral venous thrombosis (CVT) in women1 are associated with pregnancy or puerperium, and CVT accounts for up to 57% of pregnancy-related strokes.2
Women with a previous episode of noncerebral VTE have a higher risk of VTE during subsequent pregnancies than outside pregnancy.3
A systematic review4 evaluating the incidence of pregnancy-related VTE, including recurrent CVT, in women with prior CVT, revealed a paucity of studies addressing this issue. Data on the risks and benefits of antithrombotic prophylaxis in this setting are even scarcer. Selection bias is also a concern as some of the included publications were patient series conducted in referral centers, instead of follow-up studies of established cohorts. As the risk of thrombotic recurrence is probably increased in the first months after the index episode, the relatively short follow-up of most studies could be another source of bias. Individual patient data on the risk factors for the index CVT are not provided in most prior reports. A trend toward lower rates of spontaneous abortion in women on prophylaxis with heparin was found, but the numbers were too low to reach statistical significance.4
We performed a follow-up study of women at fertile age previously included in a large multicenter prospective cohort of patients with acute CVT to evaluate the outcome of their pregnancies, namely (1) CVT recurrence and occurrence of noncerebral VTE, (2) adverse outcomes of pregnancy, and (3) antithrombotic prophylaxis use and complications.
We included female patients with prior CVT at fertile age, who participated in the ISCVT (International Study on Cerebral Vein and Dural Sinus Thrombosis).5 ISCVT was a prospective multicenter cohort study conducted from 1998 to 2002 that included 624 adult patients with diagnosis of CVT confirmed following established criteria. Detailed information on demographics, risk factors (including thrombophilia), and outcome were systematically recorded. We defined fertile age as ≤45 years at the time of CVT. All participating centers were invited to participate. Ethical Committee approval was obtained in all centers according to local legislation.
Patients willing to participate were scheduled a consultation or received a phone interview. A standardized case report form for assessment of each pregnancy after CVT was used as guidance for the clinical interview. Whenever the patient reported pregnancy occurring after CVT, the following information was obtained: maternal outcomes (CVT/noncerebral VTE/nonthrombotic complications), fetal outcomes (spontaneous abortion/induced abortion/fetal death/preterm birth/full term birth/congenital anomalies), and antithrombotic prophylaxis (all trimesters of pregnancy and puerperium). Puerperium was defined as the first 6 weeks after delivery. Imaging confirmation was required for the diagnosis of recurrent CVT or other VTE.
Crude rates of recurrent CVT and noncerebral VTE (per 1000 pregnancies) and spontaneous abortion (%) were calculated. In a secondary analysis, these rates were stratified according to antithrombotic therapy. A P value of 0.05 was considered statistically significant. The 95% confidence intervals (95% CI) were calculated by the Wilson method. Analyses were conducted using SPSS statistics 20 (IBM Corp, NY).
Of the 465 female patients included in the ISCVT cohort, 314 were eligible. The 75 centers that included these patients in the original cohort were invited to participate. Thirty-two centers, corresponding to 193 of the eligible patients, agreed to participate. One hundred and nineteen patients gave informed consent. One patient refused to provide consent, and 73 patients could not be reached. The baseline characteristics of participants and nonparticipants were similar (Table I in the online-only Data Supplement). The 32 participating centers were spread across countries (Table II in the online-only Data Supplement). Most included patients lived in high-income countries (100 patients, 84%). The median follow-up was 169 months (minimum 135 months).
Eighty-two pregnancies occurred in 47 women (Figure). The mean age at the index CVT in patients after pregnancy was 23 years (SD, 5). The mean age at pregnancy was 30 years (SD, 5; minimum 19, maximum 41). Sixty percent of pregnancies occurred in women with a permanent risk factor identified at the time of the index CVT, and in all remaining cases, at least 1 transient risk factor was found. Eight pregnancies occurred in women whose index CVT occurred during pregnancy/puerperium. The distribution of risk factors is described in Table III in the online-only Data Supplement.
Rate of Recurrent CVT and Noncerebral VTE Associated With Pregnancy
There was 1 recurrent CVT among 82 pregnancies after CVT (12 per 1000 pregnancies; 95% CI, 2–66). This recurrence was observed in a woman whose index event, 31 months before, was associated with oral contraception and protein S deficiency. The recurrence was diagnosed in the first trimester (gestational age, 10 weeks), and the patient was receiving low-molecular-weight heparin (LMWH) in prophylactic dosage at that time. The pregnancy proceeded but was further complicated by intrauterine growth restriction and preterm labor. The mother received LMWH at therapeutic dosage and fully recovered in the follow-up.
Noncerebral VTE occurred in 2 pregnancies (24 per 1000 pregnancies; 95% CI, 7–85). Both women had a diagnosis of deep vein thrombosis during puerperium, together with pulmonary emboli in one of them. In the latter, deep vein thrombosis and pulmonary emboli occurred despite prophylaxis with LMWH. The index CVT occurred during the postpartum period in association with protein S deficiency. The former was not receiving any antithrombotic prophylaxis after cesarean section. Her index CVT was associated with oral contraception.
The majority of women received antithrombotic prophylaxis during pregnancy and puerperium, predominantly LMWH (66% of third trimester pregnancies and 61% of puerperal periods; Table IV in the online-only Data Supplement). The antithrombotic therapy in each pregnancy trimester and postpartum period is described in Figure I in the online-only Data Supplement.
Other Maternal Outcomes
Maternal medical complications were reported in 10 pregnancies (12%; Table V in the online-only Data Supplement). There was 1 case of maternal death because of eclampsia in a woman whose index CVT was related with oral contraceptives and was not receiving any antithrombotic prophylaxis.
Fourteen pregnancies ended in spontaneous abortion, resulting in an overall crude risk of 17% (95% CI, 11–27). Other fetal outcomes are described in Table VI in the online-only Data Supplement. Stratification of miscarriage according to antithrombotic prophylaxis (after exclusion of induced abortions) shows a significantly different rate among treatment groups: 8 of 24 pregnancies in women without prophylaxis versus 1 of 26 pregnancies in women receiving heparin; relative risk, 0.12 (P=0.03). Five of 10 women who took oral anticoagulation during the first trimester had spontaneous abortion/stillbirth. Three cases of congenital malformations (4%) were reported, all in women receiving antithrombotic prophylaxis during the first trimester.
We were able to include in this follow-up study 119 women at fertile age with prior CVT. During a median follow-up of 14 years after CVT, these patients had 82 pregnancies. VTE related with pregnancy occurred in 3 women, including 1 recurrent CVT. Although this cohort is multinational, this study mostly describes patients from high-income countries followed at tertiary centers. A large proportion of patients received antithrombotic prophylaxis, with a rate of LMWH prescription ≈60% during the third trimester and puerperium.
We found an overall rate of 12 recurrent CVT per 1000 pregnancies (95% CI, 2–66) and 24 noncerebral VTE per 1000 pregnancies (95% CI, 7–85). These results are in accordance to those described in a prior systematic review.4 This low event rate makes the current sample unpowered to detect statistically significant differences between antithrombotic prophylaxis strategies. The rate of spontaneous abortion was 17%. Again, this result is in accordance with the prior systematic review.4 A significantly lower proportion of spontaneous abortions occurred in patients receiving heparin. Although a recent systematic review suggested no clear benefit of LMWH in preventing recurrent pregnancy loss in women with thrombophilia,6 lack of statistical power because of small sample size cannot be excluded. Other studies are in accordance with our findings.7
This study has limitations, the most important being related with data collection through patient interview, susceptible to recall bias. However, this is minimized by the fact that pregnancy is an objective and definite event, a local neurologist conducted the interview, and imaging confirmation was required for the diagnosis of VTE. Another possible source of bias is the lack of information about the reasons for not becoming pregnant. Although the rate of women with after pregnancy was high (40% of women with CVT before the age of 45 years), we cannot exclude negative medical counseling on pregnancy in some cases. About one third of the eligible patients from the original cohort could not be reached. Although the characteristics of these women did not differ from those of the included patients and all patients belonged to a prior well-defined prospective cohort, death from thrombotic events could not be excluded. The long follow-up (median, 14 years) guarantees that most of the fertile period of the affected women was covered, minimizing the time bias commonly observed in studies in which the analysis is limited to pregnancies occurring shortly after the index event, probably more prone to venous thrombotic recurrence. Additional strengths of the study are standardized study of CVT pathogenesis, including search for prothrombotic factors, and detailed description of antithrombotic prophylaxis and hemorrhagic complications. Because of low feasibility of a long-term prospective study of women at fertile age with previous CVT and the ethical concerns of conducting randomized trials on antithrombotic prophylaxis in this population, the design of this study is probably the best feasible option to assess the risk of pregnancy-related venous thrombosis and overall pregnancy outcome in women with prior CVT.
The results from this study support the accumulating evidence on safety of pregnancy after CVT and provide further insights on the best preventive strategy.
This is the largest study evaluating safety of pregnancy in women with prior CVT. The most important finding is the description of a low absolute risk of recurrent venous thrombosis associated with pregnancy in women with prior CVT followed in tertiary settings. The fetal outcome is favorable, with a significantly lower rate of spontaneous abortion in women receiving antithrombotic prophylaxis with heparin in the first trimester. Our results indicate that pregnancy should not be contraindicated in women with prior CVT.
ISCVT-2 PREGNANCY Investigators (the Name of the Centers Is Given in Parenthesis):
Diana Aguiar de Sousa, Patrícia Canhão, Teresa Pinho e Melo, José Manuel Ferro (Hospital Santa Maria, Lisbon, Portugal); Liliana Pereira, Elisa Costa, Miguel Rodrigues (Hospital Garcia de Orta, Almada, Portugal); Miguel Viana Baptista (Hospital Egas Moniz, Lisbon, Portugal); Marta Carvalho (Hospital de São João, Oporto, Portugal); Gabriela Lopes, Manuel Correia (Hospital de Santo António, Oporto, Portugal); Ana Amélia Pinto (Hospital Fernando da Fonseca, Amadora, Portugal); Paulo Coelho, Luís Cunha (Centro Hospitalar da Universidade de Coimbra, Coimbra, Portugal); Andreia Veiga, Mário Silva (Hospital de São Pedro,Vila Real, Portugal); Maria Manuela Costa (Hospital Pedro Hispano, Matosinhos, Portugal); Hipólito Nzwalo, Fátima Ferreira (Hospital de Faro, Faro, Portugal); Jonathan Coutinho, Yvonne Zuurbier (Academic Medical Centre, Amsterdam, the Netherlands); Nelly Dequatre-Ponchelle, Didier Leys (Centre Hospitalier Universitaire de Lille, France); Yannick Bejot, Maurice Giroud (Hospital General, Dijon, France); José Luis Ruiz-Sandoval (Hospital Civil de Gaudalajara, Mexico); Werner Hacke, Sven Poli, Christoph Gumbinger (Universtat Heidelberg, Germany); Christian H. Nolte (Charité Universitätsmedizin Berlin, Germany); Ulrich Bogdahn, Felix Schlachetzki (Klinik für Neurologie der Universität Regensburg, Germany); Tibo Gerriets, Erwin Stolz (Universitatsklinikum Giessen, Germany); Vanessa Bastos, Rubens Gagliardi (Faculdade de Medicina da Santa Casa de São Paulo, Brazil); Exuperio Diez-Tejedor, Blanca Fuentes (Hospital Universitário La Paz, Madrid, Spain); Ignacio Casado-Naranjo (Complejo Hospitalario de Cáceres, Spain); Ida Martinelli (IRCCS Maggiore Hospital, Milan, Italy); Umberto Scoditti (Parma University, Italy); Gabrielle de Veber (Hospital for Sick Children, Canada); Sylvain Lanthier (Centre Hospitalier de l’Université de Montreal, Quebec, Canada).
Sources of Funding
This study was supported by Fundação para Ciência e Tecnologia-SFRH/SINTD/92677/2013.
Guest Editor for this article was Ajay K. Wakhloo, MD, PhD.
Presented in part at the European Stroke Conference, Barcelona, Spain, May 10–12, 2016.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.117.018829/-/DC1.
- Received August 6, 2017.
- Revision received July 22, 2017.
- Accepted August 10, 2017.
- © 2017 American Heart Association, Inc.
- Ferro JM,
- Bousser MG,
- Canhão P,
- Coutinho JM,
- Crassard I,
- Dentali F,
- et al
- Cantu-Brito C,
- Arauz A,
- Aburto Y,
- Barinagarrementeria F,
- Ruiz-Sandoval JL,
- Baizabal-Carvallo JF
- Pabinger I,
- Grafenhofer H,
- Kyrle PA,
- Quehenberger P,
- Mannhalter C,
- Lechner K,
- et al
- Aguiar de Sousa D,
- Canhão P,
- Ferro JM
- Ferro JM,
- Canhão P,
- Stam J,
- Bousser MG,
- Barinagarrementeria F
- Skeith L,
- Carrier M,
- Kaaja R,
- Martinelli I,
- Petroff D,
- Schleußner E,
- et al
- Bouvier S,
- Cochery-Nouvellon E,
- Lavigne-Lissalde G,
- Mercier E,
- Fabbro-Peray P,
- Balducchi JP,
- et al