Feske, Steven K. MD; Singhal, Aneesh B. MD
PHYSIOLOGY AND PATHOPHYSIOLOGY OF PREGNANCY IN RELATION TO VASCULAR DISEASE
Pregnancy is a state of high metabolic demand. The cardiovascular changes of pregnancy prepare the maternal circulation to meet that demand. Estrogen and other hormones cause an increase in renin activity, leading to retention of sodium and water. This supports an increase in plasma volume beginning in the first trimester around 6 weeks of gestation. Some studies suggest that this increase in plasma volume reaches a plateau in the third trimester. Others suggest a progressive increase until term.1 Red blood cell mass also increases, but proportionally less, resulting in a mild physiologic hemodilutional anemia of pregnancy. Cardiac output, stroke volume, and heart rate increase 30% to 50% as a result of the increased demand of the developing fetus and placenta and maternal hypervolemia. This begins as early as the fifth gestational week and reaches a plateau in the late second or third trimester. Heart rate continues to rise until term. Increase in prostacyclin and redistribution of high flow in the low-resistance uteroplacental circulation and breasts and kidneys cause systemic vascular resistance (SVR) to begin to fall around the fifth gestational week. This drop in SVR is accompanied by a fall in the systolic and diastolic blood pressure. It reaches a nadir in the third or late second trimester, about 20 to 32 weeks, after which it rises until term to levels at or slightly above the normal nonpregnant blood pressure. There is increased venous capacitance and reduced blood flow (ie, relative venous stasis), often accompanied by orthostatic intolerance. The volume increase tends to increase preload. However, preload is heavily dependent upon position, with reduced preload and orthostasis resulting from the higher venous capacitance and from compression on the inferior vena cava in the supine position.
Vascular and Connective Tissue Changes
Pregnancy causes remodeling of the heart and all blood vessels. The walls of systemic arteries show a reduction in collagen and elastin content and a loss of distensibility. Animal models have shown increased contractile force, decreased stiffness, and increased relaxation response in the middle cerebral artery.2 The roles of the molecular promoters underlying these physiologic responses to pregnancy and the full effects of these adaptive changes are complex and poorly understood. Their relationship to stroke is not certain; however, one might expect the hemodynamic changes near term combined with these structural changes of blood vessels to result in a state in which more vulnerable vascular walls experience greater hemodynamic stress, possibly contributing to the risk of hemorrhage in late pregnancy.
Changes in the Coagulation System
Both hemodynamic and biochemical changes make pregnancy a state of hypercoagulability. Decreased venous compliance results in venous stasis and congestion. Compression of the inferior vena cava, aorta, and blood supply of the gravid uterus can cause vascular injury. Levels of procoagulant factors, coagulation inhibitors, and other mediators of clot formation and lysis are altered by pregnancy resulting in a state of increased hypercoagulability in late pregnancy. Levels of procoagulant factors I, VII, VIII, IX, X, XII, and XIII increase during pregnancy. There is little change in levels of factors II, V, and XI. The levels of some coagulation inhibitors fall during pregnancy. The coagulation inhibitor antithrombin III falls and is at its nadir in the third trimester. Total and free levels of the coagulation inhibitor cofactor protein S are significantly decreased as well. Although levels of protein C remain unchanged, almost a third of women have functional activated protein C resistance during the third trimester. These changes of venous flow and the molecular mediators of thrombosis are greatest during the late third trimester and early puerperium. Along with iron deficiency and the acute phase responses of the trauma and hemorrhage of delivery, they create the greatest hypercoagulability during the early puerperium.
Preeclampsia is a multisystem disorder of mid- to late pregnancy traditionally characterized by pregnancy-induced hypertension, edema, and proteinuria; eclampsia is the development of seizures in a patient with preeclampsia. Preeclampsia and eclampsia can infrequently occur after childbirth, usually within 48 hours (postpartum preeclampsia), or up to 4 to 6 weeks after childbirth (late postpartum preeclampsia).3 Postpartum preeclampsia can manifest de novo postpartum or result from preeclampsia, preexisting chronic hypertension, or persistent gestational hypertension (systolic blood pressure greater than 140 mm Hg or diastolic blood pressure greater than 90 mm Hg without proteinuria).
The criteria for diagnosis depend on the definition of the clinical syndrome. Modern consensus definitions all endorse pregnancy-induced hypertension beginning after the 20th week with proteinuria as preeclampsia, and generally accept the inclusion of women with pregnancy-induced hypertension without proteinuria but with other common manifestations, including (1) cerebral symptoms, (2) epigastric or right upper quadrant pain with nausea or vomiting, or (3) thrombocytopenia and abnormal liver enzymes.4 It has been claimed that, although pregnancy-induced hypertension is the cornerstone of clinical diagnosis, a significant proportion of women who develop hemolysis, elevated liver enzymes, and low platelet count (HELLP) syndrome or eclamptic seizures do not have hypertension.5,6 Since the precise diagnosis of preeclampsia is currently limited by the lack of a specific biomarker, it is often desirable to entertain a permissively broad definition for use in clinical practice. Although much has been learned about preeclampsia over the last decade, a clear understanding of the relationship of the many underlying risk factors and various features of abnormal physiology is still lacking. There is strong evidence that immune maladaptation is central to its cause. While abnormal placentation and placental hypoperfusion play an important role, hypoperfusion does not seem to be primary, because markers of preeclampsia are present in the first trimester, before placental hypoperfusion.7 Although the triggers for the abnormal placentation are not clearly understood, there is strong evidence for dysregulation of angiogenic and vasoactive factors, such as vascular endothelial growth factor (VEGF) and placental growth factor (PlGF) and nitric oxide and, ultimately, antagonism of these factors by binding to soluble VEGF receptor-1 (also known as Fms-like tyrosine kinase 1 [sFlt1]). Fundamental to many of the relevant clinical complications of preeclampsia are endothelial dysfunction, absence of the normal stimulation of the renin-angiotensin system despite hypovolemia, hypersensitivity to angiotensin II leading to increased systemic vascular resistance and hypertension, augmentation of the normal thromboxane A2/prostacyclin ratio, increased platelet activation, and increased thrombin formation and fibrin generation. Endothelial dysfunction contributes to increased capillary permeability that underlies proteinuria, edema, and, most important among our concerns, brain edema at relatively modest elevations of blood pressure.
ISCHEMIC STROKE ASSOCIATED WITH PREGNANCY
No long-term prospective studies of the incidence and types of pregnancy-associated stroke have yet been done. Available studies handle spontaneous and therapeutic abortions and stillbirths differently and use different definitions of the puerperium and different methods of stroke classification. This lack of consistency of data limits comparisons, but we can estimate the impact of stroke from these studies. Table 4-1.8 summarizes nine major studies published since 1985.9–17 From the three population-based studies, we can estimate the incidence of stroke. In these three studies, the incidence of all types of stroke ranges from four to seven cases per 100,000 pregnancies. However, data from the National Inpatient Sampleof the Healthcare Cost and Utilization Project suggest that the incidence of pregnancy-associated stroke has risen since the 1990s. These data estimate the incidence of all types of pregnancy-associated stroke, including ischemic and hemorrhagic strokes, subarachnoid hemorrhage, and cerebral venous sinus thrombosis, to be 25 to 34 per 100,000 deliveries. This analysis compares data from 1994 and 1995 to data from 2006 and 2007. Between these intervals, antenatal hospitalizations increased by 47% (from 0.15 to 0.22 per 1000 deliveries), and postpartum hospitalizations by 83% (from 0.12 to 0.22 per 1000 deliveries), with the increases largely explained by concurrent hypertensive disorders or heart disease.18 In comparison, the incidence of stroke in nonpregnant women in the 15- to 45-year age group is 11 per 100,000.19 Kittner found noincreased risk of ischemic stroke during pregnancy (relative risk 0.7) but a large relative increase during the 6-week postpartum period (relative risk 8.7). Figure 4-1 shows a similar preponderance of events in the postpartum period in the authors’ study. The higher proportions of strokes in the other studies are probably due to referral bias, since these represent series from single large referral hospitals. Despite its low overall incidence, stroke contributes a major proportion of the long-term disability resulting from pregnancy.
Although differences and limitations in methods of case assessment compromise interpretations, the ischemic strokes can be broken down by types to assess the mechanisms and contributing causes of pregnancy-related strokes (Table 4-2.8). Here, the authors have classified venous sinus thrombosis with ischemic-thrombotic stroke, although many will have components of hemorrhage. When mechanisms of ischemic stroke are identified, the major ones are cardioembolism and venous sinus thrombosis. Preeclampsia/eclampsia appears to be a major contributor to stroke risk. In studies in which preeclampsia/eclampsia is reported, it is present in 11% to 47% of cases of stroke. The contribution of preeclampsia/eclampsia to the cause of strokes is presumed to be complex and related to the various vasculopathic and prothrombotic effects discussed above. Other well-established causes of stroke in young patients, such as arterial dissection and moyamoya syndrome, can present during pregnancy and should be considered (Case 4-1). Other pregnancy-specific causes, such as peripartum cardiomyopathy, choriocarcinoma, and embolization of amniotic fluid or air, are very rare and should be considered based on the clinical presentations. Amniotic fluid embolism should be considered when evidence of diffuse or multifocal brain ischemia is present and accompanied by features of pulmonary embolism.
A 20-year-old woman developed right hemichorea during the second trimester of her first pregnancy. Brain imaging showed subcortical infarctions predominantly in the left hemisphere and severe stenosis of the right and left middle cerebral arteries (Figure 4-2). No headache or fever or segmental arterial narrowing or other evidence of cerebral arteritis or infection was present. Hemoglobin electrophoresis was normal, ruling out sickle cell anemia.
Comment. This patient was diagnosed with moyamoya disease, an idiopathic noninflammatory cerebral arteriopathy. The choreiform movements resolved after a short course of steroids. The patient declined the option of surgical intervention with the encephaloduroarteriosynangiosis procedure, which has been shown to reduce the risk for future stroke.19,20 She was treated with aspirin 325 mg/d for stroke prevention. She went on to have three vaginal deliveries without further neurologic symptoms such as headache, chorea, or weakness. Follow-up brain imaging studies showed no evidence for new stroke.
Stroke is suspected clinically based on the sudden onset of a neurologic deficit suggestive of a focal lesion and without an alternative cause. Properly timed neuroimaging will confirm the great majority of ischemic strokes. Concerns arise when planning neuroimaging during pregnancy. These are discussed in detail in the article “Neuroradiology in Women of Childbearing Age” by Drs Riley Bove and Joshua Klein in this issue of CONTINUUM. In general, with proper precautions, CT and MRI should be used as with nonpregnant patients to identify areas of infarction and to investigate the cerebral vasculature. Cerebral vessel imaging with magnetic resonance (MR) angiography, CT angiography, or transfemoral catheter angiography is indicated to assess for cerebral arterial dissection, reversible cerebral vasoconstriction syndrome, moyamoya disease, or other arteriopathies. Cardiac ultrasound should be performed in patients with embolic infarctions, and should include agitated saline injection (bubble study) to investigate for a right-to-left shunt from a patent foramen ovale, the presence of which may suggest paradoxical embolism and warrant further testing for lower extremity or pelvic deep vein thrombus.22 As with any patient with stroke, blood tests such as lipid panel, hemoglobin A1C, erythrocyte sedimentation rate, C-reactive protein, and others should be routinely performed, with additional tests such as hemoglobin electrophoresis for sickle cell disease or antiphospholipid antibody panel performed on a case by case basis. Genetic testing for thrombophilia (prothrombin G20210A mutation, factor V Leiden mutation, methylenetetrahydrofolate reductase mutation) can be performed during pregnancy, but testing for other hypercoagulable states (protein C, protein S, and antithrombin III deficiency) should be performed atleast 6 weeks after delivery when physiologic changes due to pregnancy will have normalized.
Therapies and Outcomes
IV recombinant tissue plasminogen activator (tPA) is a US Food and Drug Administration (FDA)–approved thrombolytic drug that remains the only proven efficacious therapy for acute ischemic stroke.23,24 Many large stroke centers offer intra-arterial thrombolysis as a salvage therapy in severe stroke cases. Intra-arterial thrombolysis typically involves mechanical clot retrieval using FDA-approved devices. Recent phase III clinical trials showing no benefit of intra-arterial over IV thrombolysis25 will prompt a reassessment of the role of catheter-based therapies, but evidence exists that improved outcomes depend onearly recanalization of occluded arteries, and FDA-approved devices will continue to find use as research proceeds on this issue. Because pregnant women were excluded from all of these trials, there has been no controlled study of the use of such agents in pregnancy. Although there are case reports of successful IV and intra-arterial thrombolytic use in pregnant women,26–31 questions for clinicians remain: Should these therapies be applied in pregnant women suffering acute ischemic strokes? Are they safe and effective for the mother? Are they safe for the fetus?
IV tPA has a very short serum half-life of less than 5 minutes. However, it binds to newly formed fibrin clots, where its lytic effect lasts for many hours. It is a large molecule that does not cross the placenta in animals, and so it should not be expected to place the fetus at risk of teratogenicity. The potential risks of real concern are maternal hemorrhage, placental hemorrhage and abruption, fetal loss, and preterm delivery. Although there are theoretical reasons to question comparisons to its use in nonpregnant patients, mainly that pregnancy is a state of relative hypercoagulability characterized by decreased intrinsic tPA and increased plasminogen activator inhibitor, the ultimate effects of these changes on the clinical efficacy of tPA are speculative and unlikely to be answered by clinical trials. Therefore, more empirical clinical data must be used to estimate the risk. The authors have found reports in the literature of six women who have received IV tPA for stroke while pregnant, although many more have been treated as this has become accepted practice.26–31 Of the six women who received IV tPA, three suffered no hemorrhagic complications, one had minor hemorrhagic transformation of the cerebral infarct, and one had an intrauterine hematoma. Of these six cases, no fetal complications occurred in three, and in two cases the pregnancy was terminated allowing no further analysis. The sixth patient and fetus died as a result not of a direct effect of the systemic tPA, but from arterial dissection complicating angioplasty. Of five women treated with intra-arterial thrombolysis for acute arterial occlusion (three tPA, two urokinase), none had serious complications from the procedure; two had hemorrhagic transformation of the stroke with good neurologic outcomes, and one had a minor buttock hematoma.26,29,32 Four of the five women delivered healthy babies; one pregnancy, in which the stroke resulted from bacterial endocarditis, ended in spontaneous abortion. It should be noted that urokinase, unlike tPA, does cross the placenta. Minor hemorrhagic transformation is common after thrombolysis in general, and it does not appear to worsen outcomes. In fact, it has been associated with better outcomes, possibly because it is a marker of early recanalization. To summarize, of these 11 women who received IV or intra-arterial thrombolysis and were reported in the literature, 10 had no major complications from thrombolysis, and the patient who died had a major complicating illness; of the 11 reported fetal outcomes, seven were delivered without complications, two were terminated therapeutically, one patient with bacterial endocarditis had a spontaneous abortion, and one fetus died along with the mother.
With such limited and uncontrolled data, it is not possible to draw firm conclusions; however, it is reasonable to judge that pregnancy does not appear to present a decisive added risk to thrombolytic therapy. Therefore, thrombolysis should be considered for all potentially disabling strokes during pregnancy, and it should not be excluded based on the fact of pregnancy alone. As in all patients, the details of the case should be carefully weighed, and patients or proxy decision makers should be well informed of risks. Obstetric consultation should be sought from the outset for careful monitoring and decision making. Care in facilities with experience both in advanced stroke care and high-risk obstetrics is optimal, and the cause and mechanism of the stroke should be carefully determined to the extent possible before therapy is prescribed. For example, women with stroke as a result of amniotic fluid embolism would not benefit from tPA. Additionally, given the known risk of cerebral hemorrhage from hypertensive encephalopathy in the setting ofpreeclampsia/eclampsia, patients who have strokes complicating preeclampsia or eclampsia should not receive tPA.
PREECLAMPSIA/ECLAMPSIA, HYPERTENSIVE ENCEPHALOPATHY, AND POSTPARTUM CEREBRAL ANGIOPATHY
Preeclampsia/eclampsia contributes to cerebrovascular events in two major ways. First, as noted above, preeclampsia/eclampsia causes many pathophysiologic changes in blood vessels and the thrombotic system and in this way accounts for a large proportion of ischemic strokes in pregnancy. Second, one direct consequence of preeclampsia/eclampsia is the posterior reversible encephalopathy syndrome (PRES), a form of the syndrome of hypertensive encephalopathy characterized by reversible brain edema, often associated with elevated blood pressure, seizures, brain hemorrhage, and ischemic strokes.
Preeclampsia/Eclampsia and Posterior Reversible Encephalopathy Syndrome
Pathophysiologic mechanisms. The pathophysiology of preeclampsia/eclampsia is discussed briefly above. The features of preeclampsia/eclampsia directly relevant to PRES and probably also to eclamptic seizures are (1) an abnormal increase in vascular tone and (2) dysfunction of endothelial cells. Patients with preeclampsia/eclampsia have heightened sensitivity to mediators of vasoconstriction, such as angiotensin II. This resultant increase in vascular tone is responsible for systemic hypertension and for the vasomotor instability that underlies vasospasm. Endothelial dysfunction is in part responsible for the instability of vascular tone, and italso results in increased vascular permeability that underlies the development of edema and proteinuria that characterize preeclampsia/eclampsia. The syndrome commonly called PRES results from the development of cerebral edema. Fluid crosses from the intravascular to the interstitial space as a result both of an increase in capillary filtration pressure caused by hypertension and of loss of integrity of the blood-brain barrier caused by endothelial dysfunction. Animal models of the blood-brain barrier’s response to severe acute hypertension have shown both increased pinocytosis and flow across impaired endothelial gap junctions.33
Postpartum Cerebral Angiopathy
Postpartum angiopathy is one of several conditions included in the spectrum of the reversible cerebral vasoconstriction syndromes (RCVS).34 Postpartum angiopathy is characterized by severe headaches and reversible narrowing of intracerebral arteries, often complicated by seizures, reversible brain edema, lobar hemorrhage, convexity (nonaneurysmal) subarachnoid hemorrhage, and ischemic strokes. In the past, patients with this syndrome were often misinterpreted as having inflammatory cerebral vasculitis because the latter can also manifest with headache, stroke, and cerebral angiographic abnormalities; however, postpartum angiopathy is a noninflammatory, vasoconstrictive condition. Approximately one-third of patients with postpartum angiopathy are noted to have reversible cerebral edema and clinical features (headaches, seizures) which are very similar to patients with PRES, and more than half the patients with PRES show evidence of cerebral artery narrowing on vascular imaging.35 Hence, postpartum angiopathy and PRES are considered overlapping conditions.36Case 4-2 illustrates this overlap in a single patient. The pathophysiologic mechanisms whereby preeclampsia/eclampsia is related to PRES are probably also applicable to postpartum angiopathy.
A 36-year-old woman developed severe headaches associated with new-onset hypertension 10 days after delivery of twins by cesarean delivery. The initial brain MRI and CT examinations were normal. Headaches persisted despite antihypertensive medications. A seizure and an episode of aphasia and hemiparesis occurred. Repeat MRI on day 18 (Figure 4-3A36) showed hyperintense regions in both parietal–occipital lobes with elevated diffusion—findings consistent with vasogenic edema. These clinical-imaging features are consistent with the posterior reversible encephalopathy syndrome (PRES). Magnetic resonance (MR) angiography of the circle of Willis showed multifocal stenoses in the proximal anterior, middle, and posterior cerebral arteries. This finding is consistent with postpartum angiopathy. MRI performed on postpartum day 19 showed hyperintense lesions on FLAIR and diffusion-weighted images, a finding consistent with ischemic stroke. Despite multiple attempts to dilate the cerebral arteries with intracerebral vasodilator injections, the patient showed clinical and angiographic progression over the course of 1 week. A follow-up MR angiogram showed worsening of the multifocal cerebral arterial stenosis, and FLAIR showed bilateral cerebral infarction with edema and hemorrhage. The patient eventually died. On autopsy, the cerebral arteries were normal, with no evidence of inflammation.
Comment. This is a classic example of postpartum eclampsia with postpartum angiopathy and features of posterior reversible encephalopathy and reversible cerebral vasoconstriction syndromes,
illustrating that these causes of postnatal ischemic and hemorrhagic stroke are interrelated conditions. These syndromes are difficult to predict or prevent. Considering these as manifestations
of preeclampsia/eclampsia, the authors treat with magnesium sulfate, based on the clinical trials
discussed in the text. It is also important to control blood pressure when elevated, as with other forms of hypertensive encephalopathy. No treatment has proven efficacy for the cerebral artery narrowing of postpartum angiopathy. Calcium channel blockers are a reasonable, if untested, choice. While 90% of patients have a self-limited course and recover within days to weeks, some patients
(as in this example) may have a progressive course and even a fatal outcome.
The syndromes mentioned above—variously called eclampsia, hypertensive encephalopathy, PRES, RCVS, and postpartum angiopathy—can then be considered as various presentations of a similar fundamental underlying pathophysiology. This clinical lumping is not meant to oversimplify the complex pathophysiology of preeclampsia/eclampsia, nor to make the claim that our understanding of this disorder is well developed. For example, there may be important pathophysiologic differences in classic preeclampsia/eclampsia and postpartum syndromes that lack proteinuria or even hypertension.4 Nonetheless, it has been recognized in recent years that the basic preeclampsia/eclampsia pathophysiology may account for these late pregnancy and postpartum syndromeswhen they do not fit the traditional definition of preeclampsia/eclampsia with proteinuria and when they develop up to many weeks after delivery.4,38
Eclamptic hypertensive encephalopathy (ie, PRES) typically presents with headache, visual symptoms referable to the occipital lobes, and seizures. Eclamptic RCVS presents with thunderclap headache, seizures, and focal neurologic deficits. Imaging typically shows posterior white and often gray matter change consistent with vasogenic cerebral edema (hypodense on CT and hyperintense on T2-weighted MRI), the findings typical of PRES, or segmental narrowing and dilation of large and medium-sized cerebral arteries, the findings typical of RCVS. Both or either of these imaging patterns may be seen, and these imaging findings have limited sensitivity, so imaging may be unrevealing in otherwise clinically convincing cases. Although most lesions are limited to edema and are therefore reversible, hemorrhages and focal ischemic strokes may also occur.
Therapies and Outcomes
The goals of therapy are to control elevated blood pressure, control seizures, and minimize vasospasm and risk of secondary infarct and hemorrhage. Because the authors interpret these syndromes in association with pregnancy as manifestation of eclampsia, treatment with IV magnesium sulfate along with other therapies directed at blood pressure and seizures is recommended. Several studies have shown the superiority of magnesium sulfate over commonly used anticonvulsants for prevention and treatment of eclamptic seizures.39–41 The authors give a loading dose of 4 g to 6 g of magnesium sulfate over 20 to 30 minutes followed by continuous infusion at 2 g per hour with an additional 2 g bolus if seizures occur during this therapy. Patients should be monitored closely and magnesium sulfate stopped if deep tendon reflexes are lost, if respirations are depressed, or if urine output falls below 100 mL in 4 hours. Calcium gluconate 1 g slow IV push can be given to reverse severe toxicity. In addition, patients with eclamptic syndromes should be treated urgently with IV antihypertensive agents and with additional antiepileptic agents if needed for seizure control. Most women with preeclampsia/eclampsia are volume contracted and will benefit from volume replacement and maintenance. Although many practitioners have used calcium channel blockers and glucocorticoids in these patients, this use is not supported by clinical data. No evidence has proven that calcium channel blockers are more effective than other antihypertensive agents. In a study of 139 patients with RCVS, of which 12 were postpartum, corticosteroids were associated with a trend toward poorer outcomes.34 This series is weighted toward the RCVS presentation, so it is not representative of eclampsia in general. A third of the patients in this series presented with seizures, and a third suffered ischemic strokes. Ninety percent had good outcomes (Modified Rankin Scale 0 to 3).34 Most patients with pregnancy-associated PRES and RCVS have a self-limited clinical course with benign outcome and resolution of brain and vascular imaging abnormalities within days to weeks; however, 5% to 12% can have a fulminant course with progressive vasoconstriction, brain edema, and strokes, culminating in persistent severe neurologic deficits or death.37,42
CEREBRAL VENOUS SINUS THROMBOSIS
This review includes the cases of cerebral venous thrombosis among those of arterial ischemic infarction above in discussing the rate of ischemic stroke in pregnancy because it is often not clearly distinguished from arterial stroke in large series of pregnancy and stroke. However, cerebral venous sinus thrombosis is a disorder very different from arterial occlusion with different pathophysiology, therapy, and outcomes. Cerebral venous thrombosis accounts for 6% to 64% of all pregnancy-associated strokes in large reported series and 17% in the authors’ series.16 Venous thrombosis may present with imaging findings of thrombus within a cerebral vein or venous sinus without parenchymal changes or with evidence of cerebral edema, apparent ischemic stroke, or hemorrhage, and as a result, this disorder is classified differently by different authors.
Thrombosis in the venous circulation, including the cerebral venous sinuses and veins, is presumed to be the outcome of the underlying hypercoagulable state of pregnancy, promoted by the various pathophysiologic changes of pregnancy described above. These effects reach their peak during the early postpartum period, the time when most cases of cerebral venous thrombosis present. Figure 4-1 shows the time during pregnancy of the diagnosis of cerebral venous thrombosis in the authors’ patients.17 In addition to the known alterations in platelet function and prothrombotic and antithrombotic proteins, iron deficiency anemia and the adaptive response to the acute trauma and hemorrhage of labor and delivery may contribute to the propensity for abnormal thrombosis. This timing of risk is comparable to lower extremity deep venous thrombosis in pregnancy.43 Although often called “venous infarction,” with large collecting sinus thrombosis, the brain lesions typically begin as areas of brain edema without infarction as a result of impaired venous drainage and increased venous pressures. Ultimately, stasis of flow may cause these lesions to progress to include areas of infarction and hemorrhage. In addition, hemorrhage may extend to other compartments, including the subarachnoid, subdural, and intraventricular spaces. Because the primary process is edema, much of the visualized lesion (hypodensity on CT or hyperintensity on T2-weighted MR) is reversible with treatment, and outcomes are typically very good, much better than for comparable-sized arterial strokes.
Women with cerebral venous thrombosis may present with headaches, focal neurologic deficits, depressed level of consciousness, or seizures, and the pregnant state should greatly heighten the index of suspicion for this diagnosis. Cerebral venous sinus thrombosis can be detected on noncontrast CT as hyperdensity in the region of thrombosis or as parenchymal hypodensity from edema or infarction or hyperdensity from hemorrhage. Contrast CT may show a filling defect within the thrombosed sinus surrounded by the enhancing dura of the sinus wall (empty delta sign). Contrast CT venography may show the thrombosis as a filling defect in the region of the affected sinus. On MRI, venous sinus thrombosis can be seen directly as thrombus with signal characteristics appropriate to the timesince onset (T1-isodense and T2-hypodense when acute, with T1 turning hyperintense followed by T2 turning hyperintense so that the thrombus is bright on both T1- and T2-weighted images when in the late subacute phase). Since cerebral venous thrombosis may have been present for some time before symptoms lead to diagnosis, it is common to see later-phase characteristics upon initial diagnosis. MR venography can also show the thrombosis as a filling defect. This can be done without contrast, an advantage during pregnancy and nursing. As with arterial stroke, with proper precautions, it is possible to obtain good imaging confirmation safely during pregnancy. Cerebral cortical vein thrombosis without venous sinus thrombosis can be more difficult to confirm but is commonly visible as anexpanded tubular vein on the cortical surface, often with T2-weighted signal hyperintensity in the adjacent parenchyma.
The best available data support, if weakly, the use of anticoagulation to treat cerebral venous thrombosis, including in those patients with hemorrhagic lesions. Meta-analysis and the authors’ experience treating many such patients are consistent with this recommendation from the literature.44,45 A randomized, controlled trial by Einha¨upl was small but seemed to show a clear benefit.46 In fact, it was small because it was terminated early due to the evidence of benefit in favor of heparin anticoagulation after only 20 patients had been enrolled. A larger Dutch study of low-molecular-weight heparin was negative but showed a trend in favor of early anticoagulation.46 This study was limited by the fact that patients in both groups received warfarin anticoagulation after the first 3 weeks of the study treatment, possibly accounting to some degree for the small difference between groups. Cerebral hemorrhage occurs in nearly half of patients with cerebral venous thrombosis, so the question of the safety of anticoagulation in this subset is important. In the Einha¨upl study, three of 10 patients treated with heparin had experienced hemorrhage before treatment. None of these patients had expansion of their hemorrhage or new hemorrhage, and all recovered fully. This study also included a retrospective review of 102 patients with cerebral venous thrombosis. Among these patients, 27 of 43 with hemorrhage received full-dose heparin, while 13 received no heparin after hemorrhages were found. Those who received heparin had lower mortality (15% versus 56%). In the Dutch study as well, no worsening occurred in those receiving anticoagulation despite the presence of hemorrhage. The authors recommend full heparin anticoagulation during the acute phase of cerebral venous thrombosis whether hemorrhage is present or not, and then a period of approximately 3 to 6 months of ambulatory anticoagulation. This extended period of anticoagulation is typically accomplished with warfarin postpartum. During pregnancy, when warfarin is contraindicated, low-molecular-weight heparin is given and then held during the period of labor and delivery (Case 4-3).
A 40-year-old woman developed progressively worsening headaches and nausea in the first trimester of her third pregnancy. She had a medical history of depression and chronic hypertension; two previous pregnancies had been uneventful. Her blood pressure was 120/78 mm Hg. The neurologic and systemic examination findings were unremarkable. On brain imaging (Figure 4-4), MRI showed hyperintense signal in the region of the right transverse sinus, and magnetic resonance (MR) venogram showed absence of flow-related signal within the right transverse sinus and decreased flow-related signal within the right sigmoid sinus and internal jugular vein—results consistent with cerebral venous sinus thrombosis. Laboratory tests showed an elevated d-dimer and a low protein S level. She was treated with low-molecular-weight heparin, and the headaches resolved within 5 days. A follow-up MR venogram performed after 2 weeks showed complete recanalization of the venous sinuses. She went on to have an uncomplicated vaginal delivery. Follow-up blood tests showed normal d-dimer and protein S levels. Six weeks after delivery, low-molecular-weight heparin was discontinued, and she began treatment with aspirin.
Comment. This case illustrates the association between pregnancy and cerebral venous sinus thrombosis. Several mechanisms, including low levels of protein S as documented in this patient, contribute to a transient hypercoagulable state during pregnancy. MR venography was preferred over CT venography to avoid radiation risks during pregnancy. This patient was treated with low-molecular-weight heparin and not warfarin because warfarin is teratogenic and can cause bleeding in the fetus.
HEMORRHAGIC STROKE AND VASCULAR MALFORMATIONS
Hemorrhagic, like ischemic, stroke is uncommon during pregnancy and the puerperium. The proportion ranges from five to 35 per 100,000 in the reported series (see Table 4-1), and estimating the incidence from the three population-based series, it ranges from 0 to 6 per 100,000. However, despite the low absolute risk, pregnancy increases the risk for hemorrhagic much more than for ischemic stroke. This risk increase is substantial during pregnancy (relative risk 2.5) and very great during the early postpartum period (relative risk 28.5).9 Despite its rarity, because of the severe implications of cerebral hemorrhage, hemorrhagic stroke is also an important cause of pregnancy-related mortality. The major established causes of pregnancy-related cerebral hemorrhage are preeclampsia/eclampsia, followed by arteriovenous malformations and aneurysms. Preeclampsia/eclampsia probably contributes even a larger portion than is apparent from the series reported in Table 4-3, since it is likely that cases that present in late pregnancy or the puerperium without proteinuria are often diagnosed and classified as of unknown cause. Other potential causes, such as disseminated intravascular coagulation, have not been reported commonly in these series.
The physiologic changes of pregnancy reviewed above include expansion of blood volume, increased stroke volume and cardiac output, rise of blood pressure from its nadir in the late second or early third trimester to near or slightly above normal as term approaches, and remodeling of vascular tissue with loss of collagen and elastin content and loss of distensibility. One might expect these changes to underlie an increased risk of hemorrhage near term. The strain and trauma of labor might be expected to add to the increased risk.
The risk of aneurysmal rupture appears to increase severalfold, rising with gestational age until it peaks at 30 to 34 weeks.48 Dias and Sakhar48 reported the mortality of pregnancy-associated aneurysmal subarachnoid hemorrhage to be 35%, with a fetal mortality of 17%. If a ruptured aneurysm is left unsecured surgically, rates of recurrent hemorrhage and maternal and fetal mortality are very high. This mortality may be greatly reduced by early surgery. In one study, subarachnoid hemorrhage without early surgery resulted in a maternal mortality of 63% and fetal mortality of 27%; these mortalities were lowered to 11% and 5%, respectively, by early surgery.48 With evidence that early surgery, open or endovascular, to secure ruptured aneurysms leads to better maternal and fetal outcomes, it is recommended that therapy for women after aneurysmal rupture proceed as it does for all patients as dictated by neurosurgical principles. Therefore, if aneurysmal subarachnoid hemorrhage occurs during pregnancy, the patient should proceed to surgery immediately, if feasible. Endovascular coiling may be an alternative with proper shielding to minimize fetal radiation exposure.49 If urgent obstetric issues (such as active labor, eclampsia, or fetal distress) prevent immediate surgery, then the woman should undergo urgent cesarean delivery followed by surgical control of the aneurysm. Because of the severe morbidity and high mortality rate of subarachnoid hemorrhage and the increased risk of rupture near term, it is recommended that unruptured aneurysms at significant risk of rupture be secured before pregnancy, whenever possible. With the high rate of screening by MR angiography for headaches and other common disorders, it is not uncommon to find small, asymptomatic, unruptured aneurysms. In general, the risk of rupture depends on size and morphology. The risk is low for small, uncomplicated aneurysms. Systematically evaluated clinical experience that would dictate the best policy for management of such aneurysms is lacking; however, it is often considered prudent to deliver such women by cesarean delivery or by vaginal methods that interrupt the second stage of labor. No clear data have been published to argue against vaginal delivery for women who have surgically secured aneurysms, and most such women can be delivered vaginally with close monitoring.
Data are conflicting concerning the influence of pregnancy on arteriovenous malformations (AVMs). Hemorrhage is the most common presenting manifestation of AVM, and AVMs that present with hemorrhage are more likely to bleed again than those discovered as a result of seizures or focal neurologic deficits. For many years, practitioners followed the analysis of Robinson and colleagues, which suggested that pregnancy increased the rate of hemorrhage of AVMs.50 A later influential analysis found a background annual rate of hemorrhage of 3.5% in women with AVM and no prior hemorrhage and 5.8% in those with prior hemorrhage, with no increase conferred by pregnancy.51 However, an analysis of the risk of rupture per day found a severalfold increase in risk on the day of delivery.52,53 Also, although overall hemorrhage rates appear to be comparable to nonpregnant women, evidence suggests that when an AVM bleeds during pregnancy, the rebleeding rate is higher than in nonpregnant women. In one study of 27 women with intracerebral hemorrhage due to AVM during pregnancy who were not treated with immediate resection, seven had recurrent hemorrhage during or immediately after pregnancy. This 26% rate of recurrent hemorrhage in the first year is significantly higher than the roughly 6% rate in nonpregnant women. Well-controlled data on which to base therapeutic decisions concerning AVMs discovered during pregnancy are lacking; however, based on the above considerations, expert recommendations are that (1) if a woman with known AVM anticipates pregnancy, the AVM should be treated before pregnancy; (2) if an AVM is discovered during pregnancy and has not bled during the pregnancy, conservative observation is usually recommended, with plans to proceed to definitive treatment after delivery; (3) if an AVM bleeds during pregnancy, consideration should be given to treatment during the pregnancy, taking into account the grade of the lesion and the expected timing of benefit in lowering risk (immediate for low-grade lesions amenable to complete surgical excision or embolization but delayed by up to 1 to 3 years for higher-grade lesions requiring radiosurgery and combination therapies).54 Although no study has shown an advantage to cesarean delivery, based on the suggestion of higher rates of hemorrhage on the day of delivery, many obstetricians will favor this approach to minimize risk.
Ischemic and hemorrhagic strokes are uncommon but serious complications of late pregnancy and the puerperium, and when they occur, they confer a major risk of long-term disability or death. Knowledge of the risks of pregnancy-associated stroke and the neurologic manifestations of preeclampsia/eclampsia will support and encourage early diagnosis and optimal management decisions. With proper precautions to minimize risk to the fetus, women can generally undergo diagnostic evaluations and be treated with aggressive measures appropriate to the severity of the condition.
- Pregnancy results in increased metabolic demand, sodium and water retention, and decrease in systemic vascular resistance, leading to expansion of plasma volume; mild anemia; increased stroke volume, heart rate, and cardiac output; and decreased systolic and diastolic blood pressures.
- Changes in vascular structure and the coagulation system, although adaptive, also lead to a relative vulnerability to hemorrhage and ischemic stroke, especially during the postpartum period.
- Preeclampsia/eclampsia is a state of hypertension, endothelial and platelet dysfunction, and enhanced coagulability with many pathologic consequences.
- The risk of ischemic stroke is increased during the postpartum period.
- Although pregnancy-associated ischemic stroke is rare, it is a major contributor to long-term disability resulting from pregnancy.
- Cardioembolism, preeclampsia/eclampsia, and cerebral venous sinus thrombosis account for most pregnancy-related ischemic stroke.
- Although data on the use of thrombolytic therapies during pregnancy are scarce, limited experience suggests that these agents can be given with safety comparable to that in nonpregnant patients.
- Preeclampsia/eclampsia can lead to several cerebrovascular syndromes, including posterior reversible encephalopathy syndrome, reversible cerebral vasoconstriction syndrome, and ischemic and hemorrhagic strokes.
- The major CNS complications of preeclampsia/eclampsia are a form of hypertensive encephalopathy. They should be treated aggressively with rapid control of blood pressure and IV magnesium sulfate.
- Cerebral venous thrombosis, especially postpartum, is one of the most common cerebrovascular complications of pregnancy.
- Patients with cerebral venous sinus thrombosis benefit from anticoagulation.
- Pregnancy increases the risk of hemorrhagic stroke. This increased risk is greatest in the postpartum period.
- The major causes of pregnancy-associated hemorrhage are preeclampsia/eclampsia and cerebral vascular malformations, such as aneurysms and arteriovenous malformations.
- Aneurysmal subarachnoid hemorrhage during pregnancy confers a high risk of death to both mother and baby.
- Women with subarachnoid hemorrhage should be seen by a neurosurgeon and undergo vascular imaging to look for aneurysm, arteriovenous malformation, or other vascular lesions.
- Aneurysmal subarachnoid hemorrhage should be treated with early surgery or endovascular techniques to secure the ruptured aneurysm and minimize the risk of recurrent hemorrhage.
- Women with arteriovenous malformations should be managed in consultation with a neurosurgeon. In most cases, they can be treated with conservative observation until after delivery. However, when hemorrhage occurs during pregnancy, patients with low-grade arteriovenous malformations may benefit from early definitive surgery or endovascular embolization.
1. Silversides CK, Colman JM . Physiological changes in pregnancy. In: Oakley C, Warnes CA. , eds. Heart Disease in Pregnancy. Malden, MA: Blackwell Publishing, 2007; : 6–17.
2. Hull AD, Long DM, Longo LD, Pearce WJ . Pregnancy-induced changes in ovine cerebral arteries. Am J Physiol. 1992; 262:(1 pt 2): R137–R143.
3. Sibai BM . Etiology and management of postpartum hypertension-preeclampsia. Am J Obstet Gynecol. 2012; 206:(6): 470–475.
4. Sibai B, Dekker G, Kupferminc M . Pre-eclampsia. Lancet. 2005; 365:(9461): 785–799.
5. Sibai BM . Diagnosis, controversies, and management of the syndrome of hemolysis, elevated liver enzymes, and low platelet count. Obstet Gynecol. 2004; 103:(5 pt 1): 981–991.
6. Douglas KA, Redman CW . Eclampsia in the United Kingdom. BMJ. 1994; 309:(6966): 1395–1400.
7. Thadhani R, Mutter WP, Wolf M, et al. First trimester placental growth factor and soluble fms-like tyrosine kinase 1 and risk for preeclampsia. J Clin Endocrinol Metab. 2004; 89:(2): 770–775.
8. Feske SK . Stroke in pregnancy. Semin Neurol. 2007; 27:(5): 442–452.
9. Kittner SJ, Stern BJ, Feeser BR, et al. Pregnancy and the risk of stroke. N Engl J Med. 1996; 335:(11): 768–774.
10. Wiebers DO, Whisnant JP . The incidence of stroke among pregnant women in Rochester, Minn 1955 through 1979. JAMA. 1985; 254:(21): 3055–3057.
11. Awada A, al Rajeh S, Duarte R, Russell N . Stroke and pregnancy. Int J Gynaecol Obstet. 1995; 48:(2): 157–161.
12. Sharshar T, Lamy C, Mas JL . Incidence and causes of stroke associated with pregnancy and puerperium. A study in public hospitals of Ile de France. Stroke in Pregnancy Study Group. Stroke. 1995; 26:(6): 930–936.
13. Witlin AG, Friedman SA, Egerman RS, et al. Cerebrovascular disorders complicating pregnancy—beyond eclampsia. Am J Obstet Gynecol. 1997; 176:(6): 1139–1145.
14. Jaigobin C, Silver FL . Stroke and pregnancy. Stroke. 2000; 31:(12): 2948–2951.
15. Jeng JS, Tang SC, Yip PK . Incidence and etiologies of stroke during pregnancy and puerperium as evidenced in Taiwanese women. Cerebrovasc Dis. 2004; 18:(4): 290–295.
16. Liang CC, Chang SD, Lai SL, et al. Stroke complicating pregnancy and the puerperium. Eur J Neurol. 2006; 13:(11): 1256–1260.
17. Feske SK, Klein AM, Ferrante KL . Clinical risk factors predict pregnancy-associated strokes. Poster presented at: International Stroke Conference, February 18–19, 2009; San Diego, CA.
18. Kuklina EV, Tong X, Bansil P, et al. Trends in pregnancy hospitalizations that included a stroke in the United States from 1994 to 2007: reasons for concern? Stroke. 2011; 42:(9): 2564–2570.
19. Grear KE, Bushnell CD . Stroke and pregnancy: clinical presentation, evaluation, treatment, and epidemiology. Clin Obstet Gynecol. 2013; 56:(2): 350–359.
20. Starke RM, Komoter RJ, Hickman ZL, et al. Clinical features, surgical treatment, and long-term outcome of adult moyamoya patients. J Neurosurg. 2009; 111:(5): 936–942.
21. Guzman R, Lee M, Achrol A, et al. Clinical outcome after 450 revascularization procedures for moyamoya disease. Clinical article. J Neurosurg. 2009; 111:(5): 927–935.
22. Colletti PM, Lee KH, Elkayam U . Cardiovascular imaging of the pregnant patient. AJR Am J Roentgenol. 2013; 200:(3): 515–521.
23. Tissue plasminogen activator for acute ischemic stroke. The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. N Engl J Med. 1995; 333:(24): 1581–1587.
24. Hacke W, Kaste M, Bluhmki E, et al. Thrombolysis with alteplase 3 to 4.5 hours after acute ischemic stroke. N Engl J Med. 2008; 359:(13): 1317–1329.
25. Broderick JP, Palesch YY, Demchuk AM, et al. Endovascular therapy after intravenous t-PA versus t-PA alone for stroke. N Engl J Med. 2013; 368:(10): 893–903.
26. Johnson DM, Kramer DC, Cohen E, et al. Thrombolytic therapy for acute stroke in late pregnancy with intra-arterial recombinant tissue plasminogen activator. Stroke. 2005; 36:(6): e53–e55.
27. Wiese KM, Talkad A, Mathews M, Wang D . Intravenous recombinant tissue plasminogen activator in a pregnant woman with cardioembolic stroke. Stroke. 2006; 37:(8): 2168–2169.
28. Leonhardt G, Gaul C, Nietsch HH, et al. Thrombotic therapy in pregnancy. J Thromb Thrombolysis. 2006; 21:(3): 271–276.
29. Murugappan A, Coplin WM, Al-Sadat AN, et al. Thrombolytic therapy of acute ischemic stroke during pregnancy. Neurology. 2006; 66:(5): 768–770.
30. Del Zotto E, Giossi A, Volonghi I, et al. Ischemic stroke during pregnancy and puerperium. Stroke Res Treat. 2011; 2011: 606–780
31. Dapprich M, Boessenecker W . Fibrinolysis with alteplase in a pregnant women with stroke. Cerebrovasc Dis. 2002; 13:(4): 290
32. Elford K, Leader A, Wee R, Stys PK . Stroke in ovarian hyperstimulation syndrome in early pregnancy treated with intra-arterial rt-PA. Neurology. 2002; 59:(8): 1270–1272.
33. Feske SK . Posterior reversible encephalopathy syndrome: a review. Semin Neurol. 2011; 31:(2): 202–215.
34. Singhal AB, Hajj-Ali RA, Topcuoglu MA, et al. Reversible cerebral vasoconstriction syndromes: analysis of 139 cases. Arch Neurol. 2011; 68:(8): 1005–1012.
35. Bartynski WS, Boardman JF . Catheter angiography, MR angiography, and MR perfusion in posterior reversible encephalopathy syndrome. AJNR Am J Neuroradiol. 2008; 29:(3): 447–455.
36. Singhal AB . Postpartum angiopathy with reversible posterior leukoencephalopathy. Arch Neurol. 2004; 61:(3): 411–416.
37. Singhal AB, Kimberly WT, Schaefer PW, Hedley-Whyte ET . Case records of the Massachusetts General Hospital. Case 8-2009. A 36-year-old woman with headache, hypertension, and seizure 2 weeks post partum. N Engl J Med. 2009; 360:(11): 1126–1137.
38. Raps EC, Galetta SL, Broderick M, Atlas SW . Delayed peripartum vasculopathy: cerebral eclampsia revisited. Ann Neurol. 1993; 33:(2): 222–225.
39. Which anticonvulsant for women with eclampsia? Evidence from the Collaborative Eclampsia Trial. Lancet. 1995; 345:(8963): 1455–1463.
40. Lucas MJ, Leveno KJ, Cunningham FG . A comparison of magnesium sulfate with phenytoin for the prevention of eclampsia. N Engl J Med. 1995; 333:(4): 201–205.
41. Altman D, Carroli G, Duley L, et al. ; Magpie Trial Collaboration Group. Do women with pre-eclampsia, and their babies, benefit from magnesium sulfate? The Magpie Trial: a randomised placebo-controlled trial. Lancet. 2002; 359:(9321): 1877–1890.
42. Fugate JE, Wijdicks EF, Parisi JE, et al. Fulminant postpartum cerebral vasoconstriction syndrome. Arch Neurol. 2012; 69:(1): 111–117.
43. Ray JG, Chan WS . Deep vein thrombosis during pregnancy and the puerperium: a meta-analysis of the period of risk and the leg of presentation. Obstet Gynecol Surv. 1999; 54: 169–175.
44. Bousser MG . Cerebral venous thrombosis: diagnosis and management. J Neurol. 2000; 247:(4): 252–258.
45. Stam J, de Bruijn SFTM, de Veber G . Anticoagulation for cerebral sinus thrombosis. Cochrane Database Systemic Reviews. 2002; (4): CD002005
46. Einha¨upl KM, Villringer A, Meister W, et al. Heparin treatment in sinus venous thrombosis. Lancet. 1991; 338:(8767): 597–600.
47. de Bruijn SF, Stam J . Randomized, placebo-controlled trial of anticoagulant treatment with low-molecular-weight-heparin for cerebral sinus thrombosis. Stroke. 1999; 30:(3): 484–488.
48. Dias MS, Sekhar LN . Intracranial hemorrhage from aneurysms and arteriovenous malformations during pregnancy and the puerperium. Neurosurgery. 1990; 27:(6): 855–865.
49. Meyers PM, Halbach VV, Malek AM, et al. Endovascular treatment of cerebral artery aneurysms during pregnancy: report of three cases. AJNR Am J Neuroradiol. 2000; 21:(7): 1306–1311.
50. Robinson JL, Hall CS, Sedzimir CB . Arteriovenous malformations, aneurysms, and pregnancy. J Neurosurg. 1974; 41:(1): 63–70.
51. Horton JC, Chambers WA, Lyons SL, et al. Pregnancy and the risk of hemorrhage from cerebral arteriovenous malformations. Neurosurgery. 1990; 27:(6): 867–872.
52. Parkinson D, Bachers G . Arteriovenous malformations: summary of 100 consecutive supratentorial cases. J Neurosurg. 1980; 53:(3): 285–299.
53. Weir B, Macdonald RL . Management of intracranial aneurysms and arteriovenous malformations during pregnancy. In: Wilkins RH, Rengachary SS. , eds. Neurosurgery. New York, NY: McGraw-Hill, 1996; : 2421–2427.
54. Ogilvy CS, Stieg PE, Awad I, et al. Recommendations for the management of intracranial arteriovenous malformations: a statement for health care professionals for a special writing group of the Stroke Council, American Stroke Association. Stroke. 2001; 32:(6): 1458–1471.
Copyright © 2014 by the American Academy of Neurology.