Document Type : Original Article
1 Heart Valve Research Center, Shaheed Rajaei Cardiovascular Medical and Research Center, Tehran University of Medical Science, Tehran, Iran
2 Heart Valve Research Center, Shaheed Rajaei Cardiovascular Medical and Research Center, Tehran University of Medical Science, Tehran, Iran;Epidemiology and Reproductive Health Department, Royan Institute for R
3 Cardiac Surgery Department, Afshar Hospital, Yazd, Iran
In pregnant women with mechanical heart valves,
the frequency of valve thrombosis increases due to
pregnancy-related hyper-coagulability. Therefore,
effective anticoagulation is critical in these patients
during pregnancy. Despite numerous research, the
optimal anticoagulation therapy during pregnancy
remains a controversial issue in the field of obstetrics.
Warfarin and heparin can be used during
pregnancy as anticoagulants in pregnant women
with mechanical heart valves, however the potential
maternal and fetal side effects associated with
these medications pose challenges (
Warfarin provides effective protection against
thrombo-embolism, but its use in pregnancy is associated
with an augmented rate of abortion and
the risk of warfarin-induced embryopathy (
Unfractionated heparin (UFH) provides an alternative
therapy that avoids fetal side effects; however,
the use of UFH is associated with increased maternal
thrombo-embolic and bleeding complications
Low molecular weight heparin (LMWH) may be
more advantageous than UFH and appears a good
There are several studies documenting the impact of mechanical heart valve replacement and anticoagulation therapy on pregnancy outcomes and risks, but most are retrospective. To date, no prospective study has been performed on Iranian women in this regard. Therefore, we designed a prospective research study to determine both maternal and fetal outcomes of pregnancy in women with prosthetic heart valves over a five year period in Rajaee Heart Hospital, as one of the cardiac surgery referral centers in Iran.
In this prospective study, 49 pregnancies in 44 women with prosthetic heart valves were followed between 2002 and 2007 in the Department of Cardiac Surgery, Rajaee Heart Hospital, Tehran, Iran.
The Ethics Committee of Rajaee Heart Hospital approved this study. Participating patients signed informed consent forms. All patients were visited during the first trimester of their pregnancies. Pregnant women referred during their second or third trimesters were excluded from the study. Patients’ clinical and socioeconomic conditions determined their anticoagulant regimens. When the patient refused the recommended anticoagulation therapy, an alternative regimen was started. Thus, we had two study groups according to the anticoagulation regimen. A total of 38 patients were on warfarin (Orion Pharma, Finland) throughout their pregnancy (group A). The remaining 11 patients (group B) had iv injections of UFH (Rotex Medica, Germany) during the first trimester (6th-12th week), after which they received warfarin until the 36th week of gestation followed by heparin for the last two weeks of pregnancy. Both groups received heparin at the time of delivery. During warfarin treatment, the international normalized ratio (I.N.R.) was checked routinely and kept between 2.0 to 3.5, as needed. During heparin treatment, the activated partial thromboplastin time (aPTT) was maintained at twice the control level. All patients underwent periodic transthoracic echocardiography (TTE) in addition to transesophageal echocardiography (TEE) when needed during the follow up period. Pediatricians examined all newborns.
Spontaneous abortion was defined as fetal loss before 20 weeks of gestation. Fetal and maternal outcomes were evaluated. Fetal outcomes included abortion, live birth, intra uterine fetal death (IUFD), preterm labor, embryopathy and type of delivery; maternal outcomes included bleeding, valvular thrombosis and maternal death.
Statistical analysis was performed using SPSS version 13 software. Continuous variables were described as mean ± standard deviation (SD). Student’s t-test compared continuous variables. The Mann-Whitney rank-sum test compared medians when the normality test failed. Noncontinuous variables were compared by either the chi-square test or Fisher's exact test, as appropriate. A p value less than 0.05 was statistically significant.
The mean age of studied women at time of pregnancy
was 29.8 ± 5.3 years (30.28 ± 5.2 in group A
and 28.09 ± 3.9 in group B). There were 36 pregnancies
in women with mitral valve replacement,
8 in women with aortic valve replacement and 5
in those with both aortic and mitral valve replacement
Type of surgeries in both groups
|Variable||Group A (warfarin)||Group B(heparin)|
|28 (73.7%)||8 (72.7%)|
|5 (13.2%)||3 (27.3%)|
MVR: Mitral valve replacement
AVR: Aortic valve replacement
Overall, 59% of pregnancies resulted in live births while 36.7% aborted. As shown in table 2, group A had 22 live births (57.9%), 15 abortions (39.5%), 1 IUFD (2.6%) and a maternal death due to acute thrombosis and prosthetic valve dysfunction in the delivery room. In group B, there were 7 live births (63.6%), 3 spontaneous abortions (27.3%) and 1 IUFD (9.1%). Live birth rates were similar between the two study groups (p=0.24).
There was no warfarin embryopathy in this study. After birth, during long term follow up of mothers and their babies, no obvious malformations were found in the infants.
Nine patients had vaginal delivery while cesarean section was performed in 20 patients. Abnormal post-partum hemorrhage was not seen in any woman. Twenty out of 49 pregnancies (40.8%) resulted in either abortion or IUFD. Despite the high rate of abortion in studied women, particularly in group A, we did not investigate the causes of spontaneous abortion. Both groups A and B had one stillbirth each (p=0.25).
Three infants were born prematurely, two in group A and one in group B.
Overall the low birth weight rate was 21% in our study (6 out of 29 cases). Although this rate was higher in group A (23% versus 14%), however this difference was not statistically significant (p>0.05).
Fetal outcome in 49 pregnancies
|Pregnancy outcome||Group A (warfarin)n=38||Group B(heparin)n=11||P value|
|15 (39.5%)||3 (27.3%)||0.35|
|9 (23.7%)||3 (27.3%)|
|22 (57.9%)||7 (63.6%)||0.5|
|7 (18.4%)||2 (18.2%)|
|15 (39.5%)||5 (45.5%)|
|1 (2.6%)||1 (9.1%)||0.4|
NVD: Normal vaginal delivery
CS: Cesarean section
Prevalence of valve-related thrombosis and prosthetic valve dysfunction (PVD) was significantly higher in group B (45.4% versus 2.6%; p<0.001). Valve-related complications were seen in 13.2% of group A cases and 54.6% from group B (p=0.004; Table 3).
Maternal complications in 49 pregnancies
|Variable||Group A(warfarin)n=38||Group B(heparin)n=11||P value|
|33 (86.8%)||5 (45.4%)||0.004|
|1 (2.6%)||5 (45.4%)||0.001|
|3 (7.9%)||1 (9.2%)||0.65|
Six patients in the heparin group had valve thrombosis and underwent additional surgery. Five of the six had severe heart failure as a result of valve thrombosis during the first trimester, while one had thrombosis at the end of the third trimester and underwent surgery during the postpartum period. Only three (50%) of these pregnancies with valve thrombosis in group B resulted in live births. In group A, valve thrombosis occurred after delivary in 2 patients (5.3%) among 4 patients with valve thrombosis.
Table 4 shows the pregnancy outcome according to warfarin dosage. Most abortions occurred in the warfarin group at doses greater than 5 mg daily (p=0.016; Table 4).
Pregnancy outcome according to warfarin dosage in group A
|Outcome||Warfarin≤5 mgn=29||Warfarin>5 mgn=9||P value|
|8 (27.6%)||7 (77.8%)||0.016|
|20 (69.0%)||2 (22.2%)||0.021|
The combination of heart disease and pregnancy
can present a challenge to the physician caring for
both the mother and fetus (
Warfarin crosses the placental barrier and its use
by the mother may result in an increased incidence
of fetal mortality and morbidity. Treatment with
this agent during pregnancy is associated with a
high spontaneous abortion rate that ranges between
16.2% and 44% (
The use of heparin presents an attractive alternative
to warfarin because it does not cross the placenta
In the present study, the overall abortion rate was
39.5% in the warfarin group of which 15.8% were
therapeutic abortions. The heparin group had more
spontaneous abortions (27.3%) compared to the
warfarin group (23.7%), although this difference
was not significant. This finding contrasted studies
by Nassar et al. (
The rate of healthy babies born by these mothers
was 57.9% in group A and 63.6% in group B,
which is similar to results reported by Nassar et al.
In this study, there was no warfarin-induced fetal
malformation, and is similar to other studies by
Vural et al. (
In the present study, thromboembolic complications
were significantly more frequent in the
heparin group (45.4%) compared to the warfarin
group (2.6%; p=0. 001). Although the live birth rate
was higher in the heparin group, the use of heparin
resulted in a higher spontaneous abortion rate. It
seems the use of subcutaneous UFH during pregnancy
raises the concern of an increased incidence
of maternal risk in the form of valve thrombosis
without any significant reduction in the incidence
of spontaneous abortion compared with warfarin
Cotrufo and colleagues described the outcomes
of 71 pregnancies in women with mechanical
valve prostheses who were anticoagulated with
warfarin for the entire duration of their pregnancies.
Although there were no maternal deaths, or
thromboembolic or hemorrhagic complications,
the rates of miscarriage (32%), stillbirth (7%) and
embryopathy (6%) increased (
One of present study’s limitations is the relatively low sample size that prevented reaching a definite conclusion.
It can be suggested that if patients on warfarin switch to heparin no embryopathy occurs, however there would be a risk of maternal complications. In addition, this study confirms previous studies that low dose warfarin (≤5 mg/day) during pregnancy is almost safe, with minimal feto-maternal complications. Additional prospective multicentric studies with more cases are recomended.