Regional Anesthesia for a Parturient with Venous Sinus Thrombosis and Placental Abruption Undergoing Fractional Heparin Therapy

A 32-yr-old gravida 1 para 0 woman at 34SYMBOLweeks presented with abdominal pain and vaginal bleeding. She had a witnessed generalized seizure at home at 29 weeks of gestation, was brought to the hospital, and was diagnosed with a left transverse sinus thrombosis and left temporal lobe cortical infarct. Serial electroencephalograms showed interictal epileptiform activity, localized to the left temporal lobe. She was started on 80 mg enoxaparin subcutaneously twice a day and remained well until the day of admission, when she noted vaginal bleeding with abdominal cramping. A nonstress test was reactive, with a fetal heart rate in the 140s and good beat-to-beat variability. Membranes were intact. Intravenous access was established with two peripheral intravenous lines. In consultation with a local high-risk obstetric center, it was considered inadvisable to transport the patient because of the continued vaginal bleeding. They further advised protamine reversal for emergency surgery.

A plan of continuous monitoring next to the operating room was decided upon. At 12 h after the patient's enoxaparin dose, a planned cesarean delivery during spinal anesthesia would be performed. If clinical deterioration and nonreassuring fetal heart rate monitoring supervened, cesarean delivery would be performed with a general anesthetic.

Although there were brief episodes of late decelerations that responded to supplemental oxygen and repositioning, the majority of the time the fetal heart rate remained in the 140s with good beat-to-beat variability. Uterine resting tone was soft throughout. At 11.5 h after the enoxaparin dose, the patient reported increased abdominal discomfort, and a primary low transverse cesarean delivery was performed during spinal anesthesia using 8 mg bupivacaine (0.75%), 10 μg fentanyl, and 0.25 mg preservative-free morphine through a 25-gauge Whitacre needle. No blood was noted in the needle before medication delivery. The operative time was 24 min. Placental abruption was confirmed, and the baby was delivered with Apgar scores of 9 (1 min; color) and 10 (5 min). The pathology specimen (placenta and adherent clot) measured 18 × 16 × 5 cm. The patient was stable in the postanesthesia care unit after delivery and was discharged from there 115 min after arrival, with a pain score of 2/10. She recovered uneventfully and was discharged on postoperative day 4. Hematology and neurology consultants recommended resumption of enoxaparin for 5–6 weeks.

Venous thromboembolism is estimated to occur in 5–10:10,000 pregnancies, approximately fivefold higher compared with nonpregnant women of similar age. Pregnancy is associated with an increased concentration of clotting factors, including I (fibrinogen), V, VII, VIII, IX, X, and XII. Thrombin generation also increases. Concurrent disorders such as obesity, lupus anticoagulant, protein S and C deficiencies, antithrombin III deficiency, and dysfibrinogenemia further increase the risk of thromboembolism.

Cerebral vein thrombosis is a subgroup of thrombotic disorders. The transverse sinus has the highest frequency of thrombosis (86%). Inherited prothrombotic tendencies such as factor V Leiden mutations, protein S and C deficiencies, and antithrombin III deficiencies account for 10–15% of cases of cerebral venous sinus thrombosis.1During pregnancy, prothrombotic risk factors or a direct cause can be identified in approximately 85% of patients with sinus thrombosis. During the last trimester of pregnancy and after delivery, the risk of sinus thrombosis is increased. The frequency of peripartum and postpartum sinus thrombosis is approximately 12 cases per 100,000 deliveries. Focal seizures with or without generalized seizures occur in 47% of affected patients. More than 80% have a good neurologic outcome.2 

Hypercoagulability and abruption may indeed be a continuum because histologic examination of uteroplacental vessels and intervillous architecture in abruption typically reveals increased fibrin deposition, thrombosis, and hypoxia-associated endothelial and trophoblast changes.3With perinatal mortality rates as high as 34%, placental abruption is a critical situation.4Large retroplacental bleeding is associated with a mortality rate of 50% or greater.

Recommendations for anticoagulation after thrombotic events have included therapeutic unfractionated or low-molecular-weight heparin (LMWH) therapy throughout pregnancy, including the antepartum period until labor. Because LMWH has greater antithrombotic activity (anti–factor Xa) than anticoagulant activity (anti–factor IIa), it does not affect the activated partial thromboplastin time. Monitoring of anti–factor Xa activity is not routinely performed. Peak anti–factor Xa activity occurs within 3–5 h of administration, and 50% of the total anti–factor Xa activity disappears within 6 h. So far, there have been no studies comparing the effect of LMWH with unfractionated heparin in the treatment of sinus thrombosis. LMWH seems to be a reasonable alternative to standard heparin therapy and has several specific advantages because it permits self-administration at home, does not require physiologic monitoring, and may reduce the risk of heparin-induced thrombocytopenia and osteoporosis. In a series of women (29 antenatal and 7 postpartum) treated with LMWH for venous thromboembolism in pregnancy, 15 received regional anesthesia for delivery and were given enoxaparin in a reduced dose of 40 mg daily, without complications.5 

Given these issues and the inevitability of a surgical delivery, our concerns centered on balancing the risks and benefits of general and regional anesthesia. A general anesthetic could possibly have adverse effects on maternal intracranial pressure and volume and would likely have sedative effects on the infant if a deeper, stress-attenuating anesthetic technique were chosen. There were further considerations of impaired uterine tone and alteration of cerebral blood flow with the effects of positive-pressure ventilation. There is a small but finite increased risk of aspiration and airway difficulty with the administration of a general anesthetic in pregnancy. Monitoring of mental status would be impossible intraoperatively. Emergence possibly poses its own risks of even transient increases in intracranial pressure as well as alteration of the seizure threshold. Finally, choosing a general anesthetic prevents maternal and paternal participation in childbirth. A subarachnoid block with a small-gauge needle seemed an acceptable alternative. An epidural approach was not considered an acceptable choice. We therefore set our plans as outlined above.

There are several long-term implications for pregnant patients with a diagnosis of venous thromboembolism, including the need for prolonged heparin therapy during pregnancy and prophylaxis during subsequent pregnancies. Moreover, the optimal duration of oral anticoagulant treatment after the acute phase is unknown. With a history of thromboembolism during pregnancy, the incidence of recurrence during a subsequent pregnancy has been estimated at 4–15%. Unfortunately, without prospective studies, firm guidelines for antepartum prophylaxis are difficult to establish.6For patients with a known hypercoagulable state or a history of venous thromboembolism unrelated to pregnancy, those with a deficiency of antithrombin III have a 70% incidence of recurrence, those with a deficiency of protein C have a 33% incidence, and those with a deficiency of protein S have a 17% incidence.7–9Coumadin therapy, started after delivery, should be continued for at least 6 weeks, and longer (e.g. , 3–6 months) for previous thromboembolic events.

Few guidelines, let alone studies, exist to inform decisions about the use and timing of regional anesthesia. The Working Group on Behalf of the Obstetric Medicine Group of Australasia, in a consensus-based guideline, suggested that therapeutic subcutaneous injections of LMWH or unfractionated heparin should be ceased at least 24 and preferably 36 h before regional anesthesia (no differentiation was made between epidural or spinal techniques).10Given the clinical situation described above and our hope that a nontraumatic spinal anesthetic after a 12-h interval would strike a favorable balance for the risks involved, we preferred the recommendation from Horlocker and subsequently from the American Society for Regional Anesthesia that needle placement for a single-dose spinal anesthetic should occur at least 10–12 h after the last LMWH dose.11 

The suggestion for protamine reversal is controversial. Although standard heparin anticoagulant effects are antagonized by protamine, because of reduced protamine binding to LMWH fractions, only the anti-IIa activity of LMWH is completely reversed; anti-Xa activity is not fully neutralized. Moreover, anti-IIa and anti-Xa activity may return up to 3 h after protamine reversal, perhaps because of additional LMWH released from subcutaneous depots.11 

We emphasize that the decision to use regional anesthesia in uncommon and complex settings such as this one must be made on an individual basis, with the risks of spinal hematoma or general anesthesia judged against their benefits.