Beruflich Dokumente
Kultur Dokumente
22, 2017
2017 THE AUTHORS. PUBLISHED BY ELSEVIER ON BEHALF OF THE AMERICAN ISSN 0735-1097
ORIGINAL INVESTIGATIONS
ABSTRACT
BACKGROUND Anticoagulation for mechanical heart valves during pregnancy is essential to prevent thromboembolic
events. Each regimen has drawbacks with regard to maternal or fetal risk.
OBJECTIVES This meta-analysis sought to estimate and compare the risk of adverse maternal and fetal outcomes in
pregnant women with mechanical heart valves who received different methods of anticoagulation.
METHODS Studies were identied using a Medline search including all publications up to June 5, 2016. Study inclusion
required reporting of maternal death, thromboembolism, and valve failure, and/or fetal spontaneous abortion, death, and
congenital defects in pregnant women treated with any of the following: 1) a vitamin K antagonist (VKA) throughout
pregnancy; 2) low-molecular-weight heparin (LMWH) throughout pregnancy; 3) LMWH for the rst trimester, followed
by a VKA (LMWH and VKA); or 4) unfractionated heparin for the rst trimester, followed by a VKA (UFH and VKA).
RESULTS A total of 800 pregnancies from 18 publications were included. Composite maternal risk was lowest with VKA
(5%), compared with LMWH (16%; ratio of averaged risk [RAR]: 3.2; 95% condence interval [CI]: 1.5 to 7.5), LMWH and
VKA (16%; RAR: 3.1; 95% CI: 1.2 to 7.5), or UFH and VKA (16%; RAR: 3.1; 95% CI: 1.5 to 7.1). Composite fetal risk was
lowest with LMWH (13%; RAR: 0.3; 95% CI: 0.1 to 0.8), compared with VKA (39%), LMWH and VKA (23%), or UFH and
VKA (34%). No signicant difference in fetal risk was observed between women taking #5 mg daily warfarin and those
with an LMWH regimen (RAR: 0.9; 95% CI: 0.3 to 2.4).
CONCLUSIONS VKA treatment was associated with the lowest risk of adverse maternal outcomes, whereas the use of
LMWH throughout pregnancy was associated with the lowest risk of adverse fetal outcomes. Fetal risk was similar
between women taking #5 mg warfarin daily and women treated with LMWH. (J Am Coll Cardiol 2017;69:268191)
2017 The Authors. Published by Elsevier on behalf of the American College of Cardiology Foundation. This is an open
access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Manuscript received October 20, 2016; revised manuscript received March 26, 2017, accepted March 28, 2017.
2682 Steinberg et al. JACC VOL. 69, NO. 22, 2017
ABBREVIATIONS Vitamin K antagonists (VKAs), such as valves; if xed doses of either UFH or LMWH were
AND ACRONYMS warfarin, are effective at reducing thrombo- administered; if <5 pregnancies were reported; if
embolic events and are standard therapy for pregnancies either were not followed to term or the
CI = condence interval
anticoagulation in the absence of contraindi- reported anticoagulation regimen was initiated after
LMWH = low-molecular-
cations (2). However, several studies have the rst trimester; if results had previously been
weight heparin
demonstrated teratogenicity of warfarin dur- published; or if studies were published in a language
MHV = mechanical heart valve
ing the sixth to ninth weeks of pregnancy other than English. Studies reporting outcomes in
RAR = ratio of averaged risk
(35), and some studies have found a high individuals with mechanical tricuspid or pulmonic
UFH = unfractionated heparin
rate of fetal loss in pregnant women taking valves, in which right-sided valve dysfunction was
VKA = vitamin K antagonist
warfarin (6). Therefore, many patients and not specically reported on, were also excluded out of
physicians have been reluctant to use a VKA during concern that these valves are at a higher risk for
pregnancy, despite guidelines in the United States thrombosis and dysfunction (1115) that could skew
and Europe that recommend it for many patients the study results. Attempts were made to include all
during pregnancy (7,8). studies reporting on left-sided MHVs. Results from
studies reporting outcomes from both right- and left-
SEE PAGE 2692 sided MHVs, in which valve dysfunction was clearly
identied as right sided or left sided, were included.
There has never been a randomized trial comparing
If, however, the position of the dysfunctional valve
different anticoagulation regimens in pregnant
could not be discerned, the study was excluded. In-
women with MHVs, and current guidelines are largely
vestigators of each study that met inclusion criteria
devised on the basis of case series and expert
were contacted if additional data were required.
consensus. The most widely cited systematic review
Four anticoagulation regimens were included in
(6) was written before the use of low-molecular-
this meta-analysis: 1) VKAs continued throughout the
weight heparin (LMWH) and included many patients
entirety of pregnancy (VKA); 2) dose-adjusted LMWH
with older-generation ball-in-cage valves, thereby
for the entirety of pregnancy (LMWH); 3) dose-
limiting the relevance of this review in contemporary
adjusted LMWH for the rst trimester, followed by a
patients. Subsequent systematic reviews and meta-
VKA for the remainder (LMWH and VKA); and 4) dose-
analyses have included women on xed-dose
adjusted UFH for the rst trimester, followed by a VKA
LMWH, which is now known to be associated with
for the remainder (UFH and VKA). We also identied
catastrophic valve failure (9,10).
women who received low-dose warfarin, dened as
The goal of this meta-analysis was to estimate the
#5 mg of daily warfarin in individuals able to maintain
risk of adverse maternal and fetal outcomes among
a therapeutic international normalized ratio (INR).
different anticoagulation regimens in a contemporary
population of pregnant women with modern MHVs.
OUTCOMES AND DEFINITIONS. The primary
METHODS maternal outcome was dened as a composite of
maternal death, prosthetic valve failure, and systemic
STUDY SELECTION. Studies were identied through thromboembolism. Prosthetic valve failure was
a Medline review using the following search terms: dened as abnormal valve function leading to a
(mechanical valve OR heart disease OR valve clinically meaningful outcome, such as heart failure,
replacement OR heart valve) AND pregnancy AND arrhythmia, or reoperation. Thromboembolism was
(unfractionated heparin OR heparin OR low molecular dened as any systemic arterial thrombotic event,
weight heparin OR enoxaparin OR warfarin OR low such as stroke or transient ischemic attack. The pri-
dose warfarin OR oral anticoagulation OR Coumadin mary fetal outcome was dened as a composite of
OR coumarin OR anticoagulation). Publications up to spontaneous abortion, fetal death, and the presence
June 5, 2016, were included in the search. Each of any congenital defect. The denition of sponta-
publication was independently adjudicated by 2 of neous abortion was not uniform among the included
the authors (Z.L.S. and E.V.K.) to determine eligibility studies. Spontaneous abortion was dened as any
for inclusion in the meta-analysis. unplanned fetal loss before 20 weeks of gestation in
Study inclusion required unambiguous reporting of 16 studies (1631), 22 weeks in 1 study (32), and 24
outcomes of interest in pregnant women with MHVs weeks in 1 study (33). Fetal death was dened as any
who had anticoagulant therapy with a VKA, LMWH, or unplanned fetal loss at or after 20 weeks of gestation
unfractionated heparin (UFH). Studies were excluded in 16 studies (1631), 22 weeks in 1 study (32), and 24
in the following circumstances: if >10% of reported weeks in 1 study (33). To avoid misclassication, we
pregnancies occurred in women with ball-in-cage used the denition provided by each individual
JACC VOL. 69, NO. 22, 2017 Steinberg et al. 2683
JUNE 6, 2017:268191 Anticoagulation in Pregnant Women With Mechanical Heart Valves
was tted on the transformed risks from individual 1) Review article (86)
2) Meta-analyses/systematic review (6)
studies, with the different regimens as xed effects
(using VKA, regardless of dose, as the reference) and
Articles remaining Excluded (n = 137)
the different cohorts of women receiving a specic (n = 155) Met exclusion criteria
regimen within studies as random effects. For the
1) < 5 cases (63)
analysis, a double-arcsine variance-stabilizing trans- 2) > 10% Ball-in-cage valves (23)
formation (34,35) was applied to the calculated risks, 3) Outcomes of interest not reported (18)
4) Language other than English (15)
which allowed the inclusion of cohorts reporting zero 5) Subtherapeutic INR target, fixed dose
events without adding a continuity correction (36). LMWH, or fixed dose UFH (8)
6) Duplicate reporting of data (6)
The means of the transformed risks estimated from Included studies 7) Included TVR and/or PVR (2)
the meta-regression were then back-transformed (37) (n = 18) 8) Pregnancies not followed to term or started
to provide an average estimate of maternal and fetal following pregnancy after the first trimester (2)
INR international normalized ratio; LMWH lowmolecular-weight heparin; NR not reported; PTT partial thromboplastin time; UFH unfractionated heparin;
VKA vitamin K antagonist.
maternal and fetal outcomes with a regimen of Data on the percentage of pregnant women with
LMWH and VKA, and all 4 studies administered ball-in-cage valves were missing in 6 studies. Of
LMWH doses on the basis of antifactor Xa levels. these, only 1 study included pregnancies before 1998
Seven studies reported on maternal and fetal and accounted for 6% of the total VKA cohort (28).
outcomes with a regimen of UFH and VKA; 5 of these The majority of published outcomes for the VKA, UFH
studies reported dose-adjusted UFH on the basis of and VKA, and low-dose warfarin regimens originated
partial thromboplastin time levels, and 2 studies did from Asia, Africa, and the Middle East (85%, 90%, and
not report the method for monitoring of UFH. Four 76% of pregnancies, respectively). Most published
studies reported fetal outcomes with a low-dose outcomes for the LMWH regimen originated from
warfarin regimen. With the exception of 1 study Europe and North America (79%). Most published
(33), every patient was transitioned to UFH before outcomes for the LMWH and VKA regimen originated
delivery. from a multinational study (72%); however, the
JACC VOL. 69, NO. 22, 2017 Steinberg et al. 2685
JUNE 6, 2017:268191 Anticoagulation in Pregnant Women With Mechanical Heart Valves
VKA
Al-Lawati, 2002 42 0 [0,0] 0 (0, 8)
Srivastava, 2002 37 0 [0,0] 0 (0, 9)
Nassar, 2004 30 0 [0,0] 0 (0, 12)
Khamooshi, 2007 142 11 [5,6] 8 (4, 13)
Kim, 2007 5 0 [0,0] 0 (0, 52)
Khamoushi, 2011 38 3 [1,2] 8 (2, 21)
Basude, 2012 22 0 [0,0] 0 (0, 15)
De Santo, 2012 16 0 [0,0] 0 (0, 21)
Samiei, 2012 43 4 [1,3] 9 (3, 22)
Hassouna, 2014 55 0 [0,0] 0 (0, 6)
van Hagen, 2015 38 2 [1,1] 5 (1, 18)
Ayad, 2016 17 0 [0,0] 0 (0, 20)
LMWH
Bauersachs, 2003 7 0 [0,0] 0 (0, 41)
Abildgaard, 2009 12 2 [0,2] 17 (2, 48)
Quinn, 2009 11 1 [0,1] 9 (0, 41)
Yinon, 2009 23 1 [1,0] 4 (0, 22)
Saeed, 2011 8 0 [0,0] 0 (0, 37)
Basude, 2012 4 3 [1,2] 75 (19, 99)
De Santo, 2012 1 1 [0,1] 100 (3, 100)
van Hagen, 2015 17 2 [1,1] 12 (1, 36)
LMWH + VKA
Descarries, 2006 5 1 [0,1] 20 (1, 72)
Quinn, 2009 1 0 [0,0] 0 (0, 98)
Basude, 2012 6 1 [0,1] 17 (0, 64)
van Hagen, 2015 31 3 [0,3] 10 (2, 26)
UFH + VKA
Al-Lawati, 2002 21 2 [0,2] 10 (1, 30)
Khamooshi, 2007 54 11 [2,9] 20 (11, 34)
Kim, 2007 18 0 [0,0] 0 (0, 19)
Khamoushi, 2011 11 6 [0,6] 55 (23, 83)
Samiei, 2012 10 5 [0,5] 50 (19, 81)
van Hagen, 2015 47 1 [0,1] 2 (0, 11)
Ayad, 2016 28 4 [0,4] 14 (4, 33)
0 20 40 60 80 100
The individual and composite maternal endpoints as reported by each publication included in the meta-analysis. The forest plot represents an
averaged risk of the composite outcome, weighted by study sample size. *Values in brackets are number of deaths, number of prosthetic
valve failures or thromboembolisms. CI condence interval; VKA vitamin K antagonist; other abbreviations as in Figure 1.
2686 Steinberg et al. JACC VOL. 69, NO. 22, 2017
VKA
Al-Lawati, 2002 42 11 [11,0,0] 26 (14, 42)
Srivastava, 2002 37 5 [2,0,3] 14 (5, 29)
Nassar, 2004 30 13 [11,2,0] 43 (25, 63)
Khamooshi, 2007 142 78 [66,5,7] 55 (46, 63)
Kim, 2007 5 4 [1,2,1] 80 (28, 99)
Khamoushi, 2011 38 10 [9,1,0] 26 (13, 43)
Basude, 2012 22 17 [17,0,0] 77 (55, 92)
DeSanto, 2012 16 0 [0,0,0] 0 (0, 21)
Samiei, 2012 43 9 [8,1,0] 21 (10, 36)
Hassouna, 2014 55 6 [5,1,0] 11 (4, 22)
van Hagen, 2015 38 29 [19,6,4] 76 (60, 89)
Ayad, 2016 17 13 [11,2,0] 76 (50, 93)
LMWH
Bauersachs, 2003 7 0 [0,0,0] 0 (0, 41)
Abildgaard, 2009 12 0 [0,0,0] 0 (0, 26)
Quinn, 2009 11 1 [0,1,0] 9 (0, 41)
Yinon, 2009 23 4 [2,2,0] 17 (5, 39)
Saeed, 2010 8 0 [0,0,0] 0 (0, 37)
Basude, 2012 4 1 [1,0,0] 25 (1, 81)
DeSanto, 2012 1 0 [0,0,0] 0 (0, 98)
van Hagen, 2015 17 2 [2,0,0] 12 (1, 36)
LMWH + VKA
Descarries, 2006 5 0 [0,0,0] 0 (0, 52)
Quinn, 2009 1 0 [0,0,0] 0 (0, 98)
Basude, 2012 6 3 [3,0,0] 50 (12, 88)
van Hagen, 2015 31 0 [0,0,0] 0 (0, 11)
UFH + VKA
Al-Lawati, 2002 21 6 [3,3,0] 29 (11, 52)
Khamooshi, 2007 54 11 [8,2,1] 20 (11, 34)
Kim, 2007 18 9 [0,8,1] 50 (26, 74)
Khamoushi, 2011 11 4 [3,1,0] 36 (11, 69)
Samiei, 2012 10 4 [3,1,0] 40 (12, 74)
van Hagen, 2015 47 9 [8,0,1] 19 (9, 33)
Ayad, 2016 28 13 [5,7,1] 46 (28, 66)
0 20 40 60 80 100
The individual and composite fetal endpoints as reported by each publication included in the meta-analysis. The forest plot represents an
averaged risk of the composite outcome, weighted by study sample size. *Values in brackets are number of spontaneous abortions, number
of fetal deaths, number of births with congenital defects. Abbreviations as in Figures 1 and 2.
JACC VOL. 69, NO. 22, 2017 Steinberg et al. 2687
JUNE 6, 2017:268191 Anticoagulation in Pregnant Women With Mechanical Heart Valves
LMWH regimen (RAR: 0.4; 95% CI: 0.1 to 0.8) LMWH + VKA 0.4 [ 0.0 , 1.1 ]
(Figure 5A). No signicant difference was observed UFH + VKA 0.9 [ 0.4 , 1.5 ]
for the remaining anticoagulation regimens. In a
subgroup of individuals taking low-dose warfarin, the 0.0 1.0 2.0 3.0 4.0 5.0
Ratio of Averaged Risks
estimated averaged risk of the fetal composite
outcome was 4.8% (95% CI: 0.0 to 16.9). When the low-
dose warfarin regimen was compared with the alter-
B Fetal composite outcome
regimen. Sensitivity analysis demonstrated that the each pregnancy. Our data lack granularity with regard
conclusions of the maternal outcomes analysis to the trimester in which the majority of adverse
remained unchanged when assuming that all re- maternal events occurred, but a possible explanation
ported maternal events in the VKA group occurred in is that the risk of thrombosis is highest in the rst
individuals with left-sided MHVs and that none of the trimester of pregnancy, when VKA risk to the fetus is
reported maternal events in the group taking UFH highest. This could also explain the high incidence of
and VKA occurred in individuals with left-sided adverse maternal outcomes observed with a regimen
MHVs (Online Table 1). Similarly, the conclusions of of UFH and VKA. It is notable that although the use of
the fetal outcomes analysis remain unchanged a VKA was associated with a high incidence of spon-
regardless of how many of the 6 reported fetal events taneous abortion, the incidence of fetal death was
occurred in individuals with left-sided MHVs who quite low, a nding supporting the idea that fetal risk
had an anticoagulation regimen of UFH and VKA to VKA exposure is highest during early gestation (4).
(Online Table 2). It is also important to recognize that the relative
A signicant amount of heterogeneity was present reduction in adverse fetal outcomes with these
in the estimates of risk among studies. Heterogeneity regimens, compared with VKA, is likely to be signi-
was estimated to account for 44% (95% CI: 18% to cantly underestimated if the comparison warfarin
82%) of the variability of the averaged risk estimates group is limited to individuals who must continue
of the maternal composite outcome and for 81% (95% taking >5 mg of warfarin daily to maintain a
CI: 67% to 90%) of the total variability of the averaged therapeutic INR, although this comparison was not
risk estimates of the fetal composite outcome, as performed in the current study.
displayed by the statistic I 2 (Online Tables 3 and 4). Our results support the American College of
Cardiology and American Heart Association guide-
DISCUSSION
lines for the management of patients with valvular
heart disease, which recommend the use of low-dose
On the basis of this contemporary meta-analysis of
warfarin in women who are able to maintain thera-
800 pregnancies in women with MHVs and modern
peutic INRs (Class IIa) over the use of either rst
anticoagulation regimens, VKA is the anticoagulation
trimester LMWH or UFH use (Class IIb) (7). It is notable
regimen associated with the lowest risk of adverse
that despite ndings supportive of these guidelines,
maternal outcomes. This nding is consistent with
only 4 of the referenced publications in the valve
those of previous publications (6,9,10), despite the
guidelines were included in our analysis because of
differences in the patients studied; we included many
our stringent inclusion and exclusion criteria.
fewer women with ball-in-cage valves and included a
study group of women taking dose-adjusted LMWH. STUDY LIMITATIONS. The data obtained for this
The risk of adverse maternal outcomes with dose- meta-analysis were observational, not randomized.
adjusted LMWH throughout pregnancy is higher Therefore, direct comparisons among regimens using
than that of a VKA regimen. This is balanced by a the RARs should be regarded as exploratory and must
signicant reduction in adverse fetal outcomes. be interpreted with caution, given the possibility of
However, on comparison with a subset of individuals confounders. For example, the observed differences
who continued taking #5 mg of warfarin throughout in adverse maternal and fetal outcomes among anti-
the duration of pregnancy, no signicant difference coagulation regimens may have been inuenced by
in fetal risk was observed, thus supporting the notion the underlying cardiac disorder (e.g., rheumatic vs.
that warfarins teratogenic effects are dose depen- congenital heart disease) that led to valve replace-
dent (42). Although the risk of maternal thrombo- ment. Additionally, the data for specic anti-
embolic complications remains a concern with coagulation regimens were clustered by region, thus
LMWH, we report a lower incidence of this compli- introducing the possibility that differences in access
cation, compared with previously published meta- to health care could have inuenced outcomes. Sec-
analyses (9,10). The exclusion of xed-dose LMWH ond, there is a paucity of published data reporting
regimens likely accounts for this difference and maternal and fetal outcomes in pregnant women with
further strengthens the argument for the use of dose- modern MHVs who are receiving many of the newer
adjusted LMWH with antifactor Xa levels during anticoagulation regimens, and this scarcity increases
pregnancy. The use of a regimen of LMWH and VKA the likelihood of a type II error. This may explain the
was not observed to have a lower risk of adverse similarity in adverse fetal events between the VKA
maternal outcomes, compared with an LMWH regimen and the 2 regimens that withhold VKA for the
regimen, despite the use of a VKA for the majority of rst trimester. Third, our data lack the granularity to
JACC VOL. 69, NO. 22, 2017 Steinberg et al. 2689
JUNE 6, 2017:268191 Anticoagulation in Pregnant Women With Mechanical Heart Valves
C ENTR AL I LL U STRA T I O N Anticoagulation Strategies in Pregnant Women With Mechanical Heart Valves
Risk of either
Risk to mother Risk to baby maternal or fetal event
(% average risk) (% average risk) (% averaged risk)
0 5 10 15 20 25 30 35 40+
% average risk
This illustration depicts maternal and fetal risk with different anticoagulation regimens. Maternal risk is lowest on a vitamin K antagonist (VKA) regimen, and fetal risk is
lowest on a low-molecular-weight heparin (LMWH) regimen. The risk of either a maternal or fetal complication during pregnancy is lowest with a low-dose warfarin
regimen; however, even low-dose warfarin carries a substantial risk of a poor outcome. CI condence interval.
determine the trimester in which each maternal event The signicant heterogeneity observed in the
occurred, thereby making it difcult to assess estimated risks is likely multifactorial. As mentioned
whether individuals receiving heparin in the rst earlier, our analysis included studies from a large
trimester and a VKA in the second and third geographic distribution and included patients from
trimesters had a thrombotic event while taking regions with vastly different access to health care,
heparin or a VKA. Finally, our composite outcomes do which may have resulted in greater heterogeneity.
not include maternal hemorrhage, premature Additionally, heterogeneity was greatest for the
delivery, fetal intracranial hemorrhage, and neonatal composite fetal outcome, which was largely driven by
death because of underreporting and nonuniform the incidence of spontaneous abortions. One expla-
denitions of these adverse events. These outcomes, nation for this nding is the apparent dose-
as well as others, are important considerations when dependent effects of VKAs on the developing fetus.
determining optimal anticoagulation regimens in this It is possible that differences in the distribution of
patient population, and they highlight the impor- VKA doses among studies were responsible for
tance of prospective registries with standardized much of the observed heterogeneity. However, given
reporting, such as the ROPAC (Registry of Pregnancy that most studies included in this group did not
and Cardiac Disease) study of pregnancy in women report the distribution of VKA doses, we were unable
with MHVs (33). to verify or explicitly model this hypothesis in the
2690 Steinberg et al. JACC VOL. 69, NO. 22, 2017
meta-regression. The random effects analysis allowed development appear to be limited to early gestation,
us to account and quantify the heterogeneity among with low incidences of fetal demise and congenital
the estimates from the different studies; however, as defects at warfarin doses of #5 mg daily. The use of
a result, studies were weighted more uniformly in the antifactor Xaadjusted LMWH, either throughout
analysis. Consequently, smaller, less-precise studies pregnancy or during the rst trimester, followed by
may have had a greater impact on the study results. warfarin use for the remainder of pregnancy, is
As such, we performed an analysis using a xed- associated with higher adverse maternal outcomes as
effects model to estimate maternal and fetal risk. compared with a VKA regimen, but with lower
Similar results were obtained, with the exception that adverse fetal outcomes. However, no difference in
fetal risk in women receiving LMWH and VKA was adverse fetal outcomes was observed between
8%, one-half of what was estimated using the random individuals taking warfarin at doses #5 mg daily and
effects model. This nding has little impact on our those with an LMWH regimen. The use of a regimen
study conclusions and further supports the notion of UFH and VKA continues to demonstrate a high risk
that avoidance of VKAs in the rst trimester of preg- of adverse maternal outcomes, without a substan-
nancy improves fetal risk. tially lower risk of adverse fetal outcomes as
Despite these limitations, we believe that reporting compared with a VKA regimen. Prospective random-
these data is important to clinical practice because no ized studies and large patient registry databases are
randomized data currently exist and randomized needed to validate these observations.
trials are unlikely, given the complex ethical and
social issues involved in choosing an anticoagulation ADDRESS FOR CORRESPONDENCE: Dr. Zachary L.
regimen for pregnant women. Therefore, our inten- Steinberg, Division of Cardiology, University of
tion is to consolidate all contemporary primary data Washington, 1959 Northeast Pacic Street, Box
relating to this subject, so that practitioners and 356422, Suite AA522, Seattle, Washington 98195-6422.
patients alike gain a greater understanding of the E-mail: zsteinberg@cardiology.washington.edu.
strength of the data that are the basis for current
recommendations.
Our meta-analysis consolidates the current experi- PERSPECTIVES
ence with up-to-date anticoagulant regimens in
contemporary study patients with modern MHVs. COMPETENCY IN PATIENT CARE AND
Additionally, our study focuses on all 4 anti- PROCEDURAL SKILLS: In pregnant women with
coagulation regimens for pregnant women with MHVs prosthetic MHVs, the use of LMWH as an alternative
endorsed by the 2014 American College of Cardiology to VKA anticoagulants results in less fetal loss but a
and American Heart Association valve guidelines (7). higher incidence of maternal complications. The risk
Our ndings support the current recommendations of VKA exposure to the fetus is dose dependent, with
put forth by the guidelines (Central Illustration). relatively low rates of adverse maternal and fetal
However, our analysis excludes individuals with right- outcomes with warfarin at doses #5 mg daily.
sided MHVs, thus making the results of this study less
clinically applicable in this patient population. TRANSLATIONAL OUTLOOK: The basis of current
knowledge of maternal and fetal risks of anticoagu-
CONCLUSIONS
lation is a small number of observational studies, and
larger, prospective studies are needed to validate
VKAs represent the safest anticoagulation regimen
these ndings.
for pregnant women with mechanical aortic and/or
mitral valves. The adverse effects of VKAs on fetal
REFERENCES
1. Kujovich JL. Hormones and pregnancy: throm- (Writing Committee to Revise the 1998 Guidelines 4. Iturbe-Alessio I, Fonseca MC, Mutchinik O,
boembolic risks for women. Br J Haematol 2004; for the Management of Patients With Valvular Heart Santos MA, Zajarias A, Salazar E. Risks of antico-
126:44354. Disease): endorsed by the Society of Cardiovascular agulant therapy in pregnant women with articial
Anesthesiologists, Society for Cardiovascular Angi- heart valves. N Engl J Med 1986;315:13903.
2. Bonow RO, Carabello BA, Chatterjee K, et al.
ography and Interventions, and Society of Thoracic
2008 Focused update incorporated into the ACC/ 5. Sareli P, England MJ, Berk MR, et al.
Surgeons. J Am Coll Cardiol 2008;52:e1142.
AHA 2006 guidelines for the management of Maternal and fetal sequelae of anticoagulation
patients with valvular heart disease: a report of the 3. Hall JG, Pauli RM, Wilson KM. Maternal and during pregnancy in patients with mechanical
American College of Cardiology/American Heart fetal sequelae of anticoagulation during preg- heart valve prostheses. Am J Cardiol 1989;63:
Association Task Force on Practice Guidelines nancy. Am J Med 1980;68:12240. 14625.
JACC VOL. 69, NO. 22, 2017 Steinberg et al. 2691
JUNE 6, 2017:268191 Anticoagulation in Pregnant Women With Mechanical Heart Valves
6. Chan WS, Anand S, Ginsberg JS. Anti- retrospective observational study. BJOG 2012;119: 30. Al-Lawati AA, Venkitraman M, Al-Delaime T,
coagulation of pregnant women with mechanical 100813; discussion 10123. Valliathu J. Pregnancy and mechanical heart
heart valves: a systematic review of the literature. valves replacement: dilemma of anticoagulation.
18. De Santo LS, Romano G, Della Corte A, et al.
Arch Intern Med 2000;160:1916. Eur J Cardiothorac Surg 2002;22:2237.
Mechanical aortic valve replacement in young
7. Nishimura RA, Otto CM, Bonow RO, et al. 2014 women planning on pregnancy: maternal and fetal 31. Srivastava AK, Gupta AK, Singh AV, Husain T.
AHA/ACC guideline for the management of outcomes under low oral anticoagulation, a pilot Effect of oral anticoagulant during pregnancy with
patients with valvular heart disease: a report of observational study on a comprehensive pre- prosthetic heart valve. Asian Cardiovasc Thorac
the American College of Cardiology/American operative counseling protocol. J Am Coll Cardiol Ann 2002;10:3069.
Heart Association Task Force on Practice Guide- 2012;59:11105. 32. Ayad SW, Hassanein MM, Mohamed EA,
lines [Published correction appears in J Am Coll
19. Samiei N, Kash F, Khamoushi A, et al. Preg- Gohar AM. Maternal and fetal outcomes in preg-
Cardiol 2014;63:2489]. J Am Coll Cardiol 2014;63:
nancy outcome after mechanical mitral valve nant women with a prosthetic mechanical heart
e57185.
replacement. J Tehran Heart Cent 2012;7:11720. valve. Clin Med Insights Cardiol 2016;10:117.
8. European Society of Gynecology (ESG), Asso-
ciation for European Paediatric Cardiology (AEPC), 33. van Hagen IM, Roos-Hesselink JW, Ruys TP,
20. Khamoushi AJ, Kash F, Hosseini S, Alizadeh
German Society for Gender Medicine (DGesGM), et al., ROPAC Investigators and the EURObserva-
Ghavidel AR, Samiei N, Haddadzadeh M. Anti-
et al. ESC guidelines on the management of car- tional Research Programme (EORP) Team. Preg-
coagulation during pregnancy in women with
diovascular diseases during pregnancy: the Task nancy in women with a mechanical heart valve:
mechanical heart valves: a prospective study. Int J
Force on the Management of Cardiovascular Dis- Fertil Steril 2011;5:4751. data of the European Society of Cardiology Reg-
eases during Pregnancy of the European Society of istry of Pregnancy and Cardiac Disease (ROPAC).
21. Saeed CR, Frank JB, Pravin M, Aziz RH, Circulation 2015;132:13242.
Cardiology (ESC). Eur Heart J 2011;32:314797.
Serasheini M, Dominique TG. A prospective trial
9. James AH, Brancazio LR, Gehrig TR, Wang A, 34. Freeman MF, Tukey JW. Transformations
showing the safety of adjusted-dose enoxaparin
Ortel TL. Low-molecular-weight heparin for related to the angular and the square root. Ann
for thromboprophylaxis of pregnant women with
thromboprophylaxis in pregnant women with Math Stat 1950:60711.
mechanical prosthetic heart valves. Clin Appl
mechanical heart valves. J Matern Fetal Neonatal Thromb Hemost 2011;17:3139. 35. Barendregt JJ, Doi SA, Lee YY, Norman RE,
Med 2006;19:5439. Vos T. Meta-analysis of prevalence. J Epidemiol
22. Abildgaard U, Sandset PM, Hammerstrm J,
10. Oran B, Lee-Parritz A, Ansell J. Low molecular Community Health 2013;67:9748.
Gjestvang FT, Tveit A. Management of pregnant
weight heparin for the prophylaxis of thrombo- women with mechanical heart valve prosthesis: 36. Kuss O. Statistical methods for meta-analyses
embolism in women with prosthetic mechanical thromboprophylaxis with low molecular weight including information from studies without any
heart valves during pregnancy. Thromb Haemost heparin. Thromb Res 2009;124:2627. events: add nothing to nothing and succeed
2004;92:74751. nevertheless. Stat Med 2015;34:1097116.
23. Quinn J, Von Klemperer K, Brooks R,
11. Thorburn CW, Morgan JJ, Shanahan MX, 37. Miller JJ. The inverse of the Freeman-Tukey
Peebles D, Walker F, Cohen H. Use of high
Chang VP. Long-term results of tricuspid valve double arcsine transformation. Am Stat 1978;32:138.
intensity adjusted dose low molecular weight
replacement and the problem of prosthetic valve
heparin in women with mechanical heart valves 38. R Core Team. R: A Language and Environment
thrombosis. Am J Cardiol 1983;51:112832.
during pregnancy: a single-center experience. for Statistical Computing. Vienna, Austria: R
12. Pterffy A, Henze A, Savidge GF, Landou C, Haematologica 2009;94:160812. Foundation for Statistical Computing, 2014.
Bjrk VO. Late thrombotic malfunction of the Available at: http://www.R-project.org/. Accessed
24. Yinon Y, Siu SC, Warshafsky C, et al. Use of low
Bjrk-Shiley tilting disc valve in the tricuspid po- March 29, 2017.
molecular weight heparin in pregnant women with
sition: principles for recognition and management.
mechanical heart valves. Am J Cardiol 2009;104: 39. Viechtbauer W. Conducting meta-analyses in R
Scand J Thorac Cardiovasc Surg 1980;14:3341.
125963. with the metafor package. J Stat Softw 2010;36:
13. Ilbawi MN, Lockhart CG, Idriss FS, et al. Expe- 148.
25. Khamooshi AJ, Kash F, Hoseini S,
rience with St. Jude Medical valve prosthesis in
Tabatabaei MB, Javadpour H, Noohi F. 40. Plesinac SD, Darko PV, Pilic IZ, Babovic IR.
children: a word of caution regarding right-sided
Anticoagulation for prosthetic heart valves in Anticoagulation therapy during pregnancy of pa-
placement. J Thorac Cardiovasc Surg 1987;93:
pregnancy: is there an answer? Asian Cardiovasc tients with articial heart valves: fetomaternal
739.
Thorac Ann 2007;15:4936. outcome. Arch Gynecol Obstet 2006;274:1415.
14. Liu P, Qiao WH, Sun FQ, et al. Should a
26. Kim KH, Jeong DS, Ahn H. Anticoagulation in 41. Kawamata K, Neki R, Yamanaka K, et al. Risks
mechanical or biological prosthesis be used for a
pregnant women with a bileaet mechanical car- and pregnancy outcome in women with prosthetic
tricuspid valve replacement? A meta-analysis.
diac valve replacement. Heart Surg Forum 2007; mechanical heart valve replacement. Circ J 2007;
J Cardiac Surg 2016;31:294302.
10:E26770. 71:2113.
15. Miyamura H, Kanazawa H, Hayashi J, Eguchi S.
42. Vitale N, De Feo M, De Santo LS, Pollice A,
Thrombosed St. Jude Medical valve prosthesis in 27. Descarries LM, Leduc L, Khairy P, Mercier LA.
Tedesco N, Cotrufo M. Dose-dependent fetal
the right side of the heart in patients with tetral- Low-molecular-weight heparin in pregnant
complications of warfarin in pregnant women with
ogy of Fallot. J Thorac Cardiovasc Surg 1987;94: women with prosthetic heart valves. J Heart Valve
mechanical heart valves. J Am Coll Cardiol 1999;
14850. Dis 2006;15:67985.
33:163741.
16. Hassouna A, Allam H. Limited dose warfarin 28. Nassar AH, Hobeika EM, Abd Essamad HM,
throughout pregnancy in patients with mechanical Taher A, Khalil AM, Usta IM. Pregnancy outcome in
heart valve prosthesis: a meta-analysis. Interact women with prosthetic heart valves. Am J Obstet
KEY WORDS fetal risk, maternal risk,
Cardiovasc Thorac Surg 2014;18:797806. Gynecol 2004;191:100913.
pregnancy
17. Basude S, Hein C, Curtis SL, Clark A, Trinder J. 29. Bauersachs R, Lindhoff-Last E. Anti-
Low-molecular-weight heparin or warfarin for coagulation of pregnant women with mechanical
anticoagulation in pregnant women with heart valves using low-molecular-weight heparin. A PP END IX For supplemental tables, please
mechanical heart valves: what are the risks? A Arch Intern Med 2003;163:27889. see the online version of this article.