Keywords
anticoagulation - pregnancy - prosthetic heart valve - heparin
Prosthetic heart valves in pregnancy present a unique clinical challenge. Adequate
anticoagulation is mandatory in such patients to prevent catastrophic complication
of valve thrombosis with up to 40% risk of maternal mortality.[1] Other less feared complications include hemorrhage and/or endocarditis. The goal
in pregnancy is to maintain adequate maternal anticoagulation while minimizing the
deleterious effects of anticoagulants in the developing fetus. Overall, the risk for
anticoagulation-related complications is ~3% per patient year.[2] We present a patient with a mechanical prosthetic mitral valve who developed intracranial
hemorrhage while on intravenous unfractionated heparin (UFH) infusion in the third
trimester.
Case Report
A primigravida was referred to our University Medical Center with a history of prosthetic
mitral valve replacement (St. Jude No. 32, St. Paul, MN) 2 years prior to this pregnancy
due to severe rheumatic mitral stenosis. The patient was otherwise healthy with no
chronic medical conditions. The patient was being treated with warfarin anticoagulation
prior to pregnancy without sequelae. As soon as the pregnancy was confirmed, the patient
was switched to enoxaparin (1 mg/kg subcutaneously twice daily) and was followed with
serial enoxaparin levels. At 14 weeks' gestation, she was converted to daily warfarin
therapy (dose range between 5 and 7.5 mg) to maintain international normalized ratio
(INR) between 2.5 and 3.5. The patient had an uneventful prenatal course.
Warfarin dose was held for 2 days and the patient was admitted at 36 weeks for conversion
to UFH infusion in anticipation for regional anesthesia and vaginal delivery. She
received a 5000 U UFH intravenous bolus followed by an infusion at a rate of 1000
U per hour. Activated partial thromboplastin time (aPTT) was monitored and infusion
adjusted to maintain aPTT between 55 and 80 seconds. Within 24 hours of initiation
of UFH infusion, the patient complained of a persistent headache that was refractory
to pain medications. The patient was afebrile and hemodynamically stable with a blood
pressure of 110/70 mm Hg. Given the clinical suspicion for an intracranial pathology,
a cranial computed tomography (CT) scan was obtained that revealed an acute subdural
hemorrhage with a 6-mm left-to-right midline shift. The patient's aPTT at the time
of diagnosis was 49 seconds (normal ≤35 seconds), platelet count 162,000, and hemoglobin
of 11.4 mg/dL. Of note, aPTT of 120 seconds was observed 12 hours prior to the onset
of headache that was managed in the standard way by holding off the infusion until
the aPTT normalized over a period of 2 hours. Heparin infusion was immediately discontinued
and consultations with neurosurgery and cardiology were obtained. The aPTT normalized
within 3 hours of intravenous heparin cessation. The patient remained hemodynamically
stable and alert and oriented without neurological deficits. She was transferred to
the neurological intensive care unit for close neurological monitoring. Neurosurgery
team obtained magnetic resonance imaging (MRI) and magnetic resonance angiography
(MRA) of the brain to delineate the source of bleeding and to rule out potential vasculature
malformations. Both MRI and MRA confirmed a stable left extra-axial hemorrhage localized
to the subdural and subarachnoid areas without evidence of abnormal brain vasculature.
The patient was expectantly managed without anticoagulation with plans to intervene
surgically in the event of clinical deterioration.
A multidisciplinary discussion was held between maternal-fetal medicine, neurosurgery,
and cardiology teams to outline further management including delivery plan. Given
her stable neurological status and a relatively late preterm gestational age of 36
weeks, the decision was made to proceed with cesarean delivery. We performed an uncomplicated
primary low transverse cesarean section under general anesthesia, delivering a vigorous
female infant with Apgar scores of 8 at 1 minute and 9 at 5 minutes and weighing 2669
g. The infant was transferred to the newborn nursery.
Following delivery, the patient returned to the neurological intensive care unit for
telemetry and serial neurological evaluations. In addition, the patient was followed
closely with serial cardiac evaluations to confirm normal function of the prosthetic
valve. The patient remained clinically stable and neurologically intact, and the headache
gradually resolved over the next 12 hours. A repeat CT scan remained stable and no
cardiac complications were encountered in the postpartum period. UFH infusion was
restarted without a loading dose ~72 hours following her delivery to aPTT level between
55 and 65 seconds (low therapeutic range). Once therapeutic on intravenous UFH, she
was started on low-dose warfarin. On hospital day 16, the patient was therapeutic
on warfarin (INR 3.0), and the UFH drip was discontinued. The patient was discharged
to home in stable condition. Subsequent serial CT scans demonstrated gradual but complete
resolution of the hematoma over the next couple of weeks.
Discussion
There are no reported cases in the literature of this devastating complication of
anticoagulation in pregnancy to our knowledge. To maintain a sustained and stable
anticoagulation profile in a pregnant woman may be challenging due to the inherent
physiological changes in coagulation profile. Prosthetic heart valves in pregnancy
are particularly at high risk of complications such as valve thrombosis, thromboembolism,
and bleeding. The estimated risk of thromboembolic complication in patients with prosthetic
heart valves in pregnancy ranges from 7 to 23%; Mortality rate may be as high as 40%
in a setting of valve thrombosis.[1]
The baseline risk for thrombosis without anticoagulation in a patient with prosthetic
mechanical heart valves is 4% per patient year, and therefore, use of anticoagulants
is mandatory. Agents used for anticoagulation may pose significant maternal and fetal
risks. Currently, the commonly used drugs in such patients include warfarin, UFH,
and low-molecular-weight heparin (LMWH). Warfarin is considered an ideal agent in
nonpregnant patients; however, it is teratogenic. Warfarin is associated with embryopathy
if used in the first trimester, spontaneous abortion, and fetal intracranial hemorrhage.
It may also restrict the use of regional anesthesia at the time of delivery. UFH is
a reliable drug when given intravenously and minimally crosses the placenta; however
maintaining adequate anticoagulation with subcutaneous administration is difficult
and is associated with higher maternal risk for thromboembolic complications.[3] Last, LMWH has a better pharmacokinetic profile and may be considered an option
for anticoagulation in pregnancy as suggested by the American Heart Association/American
College of Cardiology guidelines.[4]
[5]
Intracranial bleeding is a challenge in the presence of prosthetic heart valve as
there is an increased risk of thrombotic complications in pregnancy. The immediate
treatment of the subdural hemorrhage warrants cessation of anticoagulant therapy.
We considered reversal of heparin anticoagulation with protamine sulfate; however,
we were concerned about the potential risk of rebound hypercoagulability leading to
valvular thrombosis, heart failure, thromboembolism, and even maternal mortality.[6] The options for management of valve thrombosis include clot lysis with thrombolytic
agents or emergency valvular surgery.[7] However, in the setting of intracranial hemorrhage, thrombolytic therapy is absolutely
contraindicated and the surgical option would have required intravenous UFH infusion
during and following surgery that could be fatal in a setting of intracranial bleeding.
And therefore, the team decided against that option.
There are a few reports in the literature regarding the management of intracranial
hemorrhage and prosthetic heart valves in nonpregnant patients on warfarin. Phan and
colleagues retrospectively analyzed data on 141 patients with “high-embolic risk”
over a 23-year period. Of them, 52 had prosthetic heart valves. In this group, discontinuation
of warfarin therapy for 1 to 2 weeks had a relatively low probability of embolic events.[8] Bertram et al validated this seemingly “safe” window of 1 to 2 weeks of discontinuation
of anticoagulation following an intracranial bleed. However, in the same report, the
authors observed a higher rate of rebleeding nearing 20% when therapeutic values were
achieved. Interestingly, 20% of patients also had evidence of embolism recurrence
if therapeutic levels were not attained, emphasizing the importance of timing of reinitiation
of anticoagulation as well as targeting the appropriate level of anticoagulation so
as to avoid clinical mishap.[9] Unfortunately, none of these series included gravid patients who are inherently
more hypercoagulable than their nonpregnant counterparts.
In conclusion, we conducted a Medline and PubMed search from 1980 to 2010 with the
search terms “prosthetic valves,” “intracranial hemorrhage,” and “pregnancy” and found
no case reports. Our goal is to raise awareness and potential management strategy
of this relatively rare adverse effect of anticoagulation in pregnancy. Such cases
mandate a coordinated multidisciplinary approach for a favorable maternal and fetal
outcome.
