MATERNAL STROKE: AN UPDATE

Abstract

Purpose of the review:

Maternal morbidity and mortality is rising in the United States, and maternal stroke is a major contributor. Here, we review the epidemiology, risk factors and current recommendations for diagnosis and acute treatment of ischemic and hemorrhagic stroke during pregnancy and postpartum, focusing on recent evidence.

Recent findings:

The incidence of maternal stroke has risen in recent years, possibly due to increasing rates of hypertensive disorders of pregnancy. The risk of maternal stroke is highest in the peripartum and early postpartum period. Preeclampsia is highly associated with reversible cerebral vasoconstriction syndrome and posterior reversible encephalopathy syndrome, and is also associated with long-term increased risk of stroke and vascular dementia. Hypertensive disorders of pregnancy, migraine, and infections are risk factors for maternal stroke. Limited data suggest that thrombolytics and endovascular reperfusion therapy are safe and effective in pregnant women with ischemic stroke, but few data are available regarding safety of thrombolytics in the postpartum period. New consensus guidelines are now available to assist with management of ischemic and hemorrhagic stroke in pregnancy.

Summary:

Many gaps remain in our understanding of maternal stroke. While risk factors have been identified, there are no prediction tools to help identify which women might be at highest risk for postpartum stroke and require closer monitoring. The risk of recurrent maternal stroke has not been adequately quantified, limiting clinicians’ ability to counsel patients. The complex pathophysiology of preeclampsia and its effects on the cerebral vasculature require further targeted study. An increased focus on the prevention, recognition, and optimal treatment of maternal stroke will be critical to reducing maternal morbidity and mortality.

Introduction

Maternal stroke, defined as an acute ischemic or hemorrhagic cerebrovascular event during pregnancy or the postpartum period, complicates approximately 30 per 100,000 pregnancies in the United States (US), a 3-fold higher risk compared to overall stroke rates in young adults (1). Maternal morbidity and mortality is increasing in the United States (US), an alarming trend which has garnered attention in both the medical and lay literature (2,3). Stroke represents a major cause of maternal morbidity and mortality, accounting for 7.4% of maternal deaths in the US from 2011–2014 (4). Epidemiological data suggest that the incidence of pregnancy-related stroke is on the rise in the US, with a 47% increase in stroke during hospitalizations for delivery from 1994–1995 to 2006–2007 and an 83% increase in postpartum stroke admissions (5). A recent population-based Canadian study showed similar trends, with maternal stroke incidence increasing from 10.8 per 100,000 deliveries in 2003–2004 to 16.6 per 100,000 in 2015–2016 (p=0.02) (6). Despite stroke being far more common in elderly individuals, strokes in young adults (18–44 year olds) account for 5% of all stroke admissions in the US (7). In a cross-sectional study using New York State administrative data, 1 in 5 strokes in women aged 18–35 was associated with pregnancy (8). In Japan, stroke is now the second leading cause of maternal mortality, with 90% of fatal maternal strokes being hemorrhagic and a substantial proportion of deaths deemed preventable upon review (9,10). The economic impact of maternal stroke has not been reported, but is likely to be sizeable, since young adults with stroke have higher costs associated with disability and loss of productivity over a longer time period than older adults with stroke (11).

Clinical data are limited to guide preventive efforts and management of maternal stroke, due to the rarity of the events, the lack of prospective studies, and the exclusion of pregnant or postpartum patients from clinical stroke trials. Nevertheless, observational studies and translational work, for example using animal models of preeclampsia, offer insights into the unique factors contributing to the pathophysiology of maternal ischemic and hemorrhagic stroke and may help to inform acute management.

In this review, we highlight recent evidence regarding the epidemiology, pathophysiology, and acute management of maternal ischemic and hemorrhagic stroke, and discuss gaps in knowledge and future directions for research.

Human and Animal Rights

All reported studies/experiments with human or animal subjects performed by the authors have been previously published and complied with all applicable ethical standards (including the Helsinki declaration and its amendments, institutional/national research committee standards, and international/national/institutional guidelines).

Epidemiology and timing of maternal stroke

Multiple observational studies have demonstrated an increased risk of stroke during pregnancy and the postpartum period. However, comparison between studies is limited due to variability in definitions of stroke, categorization of timing (antepartum, peripartum, or postpartum), and study design (use of administrative data (5,6,12,13) versus registry-based (14–16) or institution-based (17–19)). While the pregnant state is widely thought of as high risk for stroke due to hypercoagulability and hemodynamic changes, the risk at time of delivery and during the puerperium appears far greater than the antepartum risk (20–23). A recent populationbased cohort study in England showed risk of first ischemic or hemorrhagic stroke was actually decreased in the antepartum period (incidence rate ratio [IRR] 0.6, 95% confidence interval [95%CI] 0.5–0.9), compared with non-pregnant women of childbearing age (16). However, a 9-fold higher relative risk of stroke was seen in peripartum women, defined as two days prior to one day after delivery (IRR 9.4, 95%CI 4.7–18.8) and the risk remained increased up to six weeks postpartum (IRR 2.7, 95%CI 1.8–4.1). Similarly, a recent study showed postpartum women to have increased risk of cerebral venous thrombosis (CVT) (adjusted odds ratio [OR] 18.7, 95%CI 8.3–41.9), but no increased risk in the antenatal period (adjusted OR 1.2, 95%CI 0.6–2.3) (24). Several other studies have demonstrated the highest risk time point to be postpartum (6,25–28) (28). Most postpartum strokes occur within the first two weeks after delivery, with 50% of readmissions for postpartum stroke occurring within 8 days after delivery (13).The risk of thromboembolic events of all types, including arterial ischemic stroke and venous infarction, is highest in the first two weeks after delivery, and contrary to previous reports which suggested that the risk was limited to the 6 weeks following the delivery, data from Kamel et al, demonstrated that the risk remains elevated up to 12 weeks postpartum (OR 10.8 95% CI, 7.8 – 15.1). (29).

Unlike overall stroke where approximately 87% of strokes are ischemic and 13% hemorrhagic (30), up to 66% of maternal strokes are hemorrhagic, including intracerebral hemorrhage (ICH) and subarachnoid hemorrhage (SAH) (1). Risk of spontaneous SAH during pregnancy increased in the US from 2002–2014, with the greatest increase seen in African American women (31). Timing of maternal hemorrhagic stroke differs depending on stroke mechanism; hemorrhagic strokes related to preeclampsia occur almost exclusively in the peripartum and postpartum period, whereas hemorrhagic strokes related to rupture of brain arteriovenous malformations occur most often in the second trimester (25,27).

Risk factors.

Identified risk factors for maternal stroke include: older age; African American race; hypertensive disorders of pregnancy (HDP) including gestational hypertension, preeclampsia and eclampsia; preexisting chronic hypertension; preexisting valvular, congenital or ischemic heart disease; sickle cell disease; cesarean section; infections; migraine; and prothrombotic states (12,32–36) (Table 1). We will highlight recent evidence concerning three of these factors: HDP, migraine, and infections.

Table 1.

Factors associated with an increased risk of stroke during pregnancy and puerperium

	Ischemic stroke		Hemorrhagic stroke
	Physiological	Pathological	Physiological	Pathological
Cardiovascular	•Venous stasis/vasodilation •Venous congestion •Direct compression from gravid uterus	•Chronic hypertension •Gestational hypertension •Preeclampsia/ eclampsia •Congenital heart disease •Rheumatic (valvular) heart disease •Peripartum cardiomyopathy •Patent foramen ovale	•Hypervolemia •Increased cardiac output	•Chronic hypertension •Gestational hypertension •Preeclampsia/eclampsia
Hematologic	•Increase in factors VII, IX, X, XII, XIII; fibrinogen; von Willebrand factor •Reduced protein S activity •Acquired protein C resistance •Antifibrinolytics produced by placenta (PAI-1, PAI-2)	•Sickle cell disease •Thrombophilias (e.g. Factor V Leiden, prothrombin 20210A mutation) •HELLP syndrome	•Anemia/hyperdynamic state	•Disseminated intravascular coagulation •Coagulopathies •Thrombocytopenia •HELLP syndrome
Neurologic		•Migraine with aura •Reversible cerebral vasoconstriction syndrome	•Capillary proliferation •Arteriolar remodeling	•Brain arteriovenous malformations •Cerebral aneurysms •Moyamoya vasculopathy •Posterior reversible encephalopathy syndrome •Migraine with aura
Obstetric	•Multifetal gestation •Multiparity •Endothelial damage due to delivery	•Cesarean section •Hyperemesis gravidarum •Amniotic fluid embolism •CVT due to low intracranial pressure from dural puncture		•Choriocarcinoma •Preeclampsia/eclampsia
Infectious/ inflammatory	•Increased activation of innate immune system •Shift from Th1 to Th2 •Heightened inflammatory response in third trimester	•Systemic lupus erythematosus •Connective tissue disorders •Antiphospholipid antibody syndrome •Puerperal infections •HIV infection Sepsis		•Sepsis
Endocrine	•Increased estrogen •Increased progesterone	•Gestational diabetes •Diabetes mellitus •Obesity

Risk factors and physiological/pathophysiological changes associated with maternal stroke. Ischemic stroke includes arterial and venous infarctions. Hemorrhagic stroke includes intracerebral hemorrhage and subarachnoid hemorrhage. CVT: Cerebral venous thrombosis. HELLP: hemolysis, elevated liver enzymes, low platelets. HIV: Human immunodeficiency virus. PAI: plasminogen activator inhibitor.

Hypertensive disorders of pregnancy.

HDP, particularly preeclampsia and eclampsia, are the most important risk factors for pregnancy-associated strokes, accounting for 25%−57% of maternal strokes (37–39). The increasing incidence of maternal stroke has been directly attributed to the increasing incidence of HDP (5). Data from the US National Inpatient Sample showed that maternal strokes in women with HDP doubled from 1994 to 2011, with 40% of the events in the HDP group occurring during the postpartum period; women with HDP were more than 5 times more likely to have a pregnancy-related stroke, compared to women without these disorders (12). In a Taiwanese study, women with preeclampsia or eclampsia had a 10-fold increased risk of hemorrhagic stroke and a 40-fold risk of ischemic stroke within 3 months postpartum (40).

The association of preeclampsia and other HDP with cerebrovascular disease does not end with the puerperium: women with HDP also have 2–3 times the risk of stroke and more than triple the risk of cerebrovascular dementia later in life (41–44). Whether women with a history of preeclampsia and other HDP should be treated with medications for primary prevention of stroke is a subject of ongoing research (45).

Migraine.

Migraine, particularly migraine with aura, is strongly associated with maternal stroke, with a systematic review finding between a 7.9 to 30.7-fold increased odds of ischemic stroke, and a 2.6 to 9.1-fold increased odds of ICH, in pregnant migraineurs (46). Proposed mechanisms include increased cerebral vasoreactivity, endothelial dysfunction and platelet activation, and a strong association between migraine and preeclampsia, which may either confound or mediate the relationship between migraine and maternal stroke. The higher prevalence of patent foramen ovale (PFO) in migraineurs may also contribute to an increased risk of ischemic stroke due to paradoxical embolism. A recent single-center study found 79% of patients aged 18–60 with cryptogenic stroke and migraine were found to have a PFO; the prevalence was 93% in patients who had frequent visual aura (47).

Infections.

Recent studies suggest that infections have a positive association with ischemic stroke, particularly in young people (48–50). Multiple mechanisms have been proposed, ranging from direct vessel wall inflammation resulting in arteritis, endocarditis complicated by septic emboli, dehydration, and cardiac arrhythmias (51). Of note, the obstetric literature has reported an association between pathogens including chlamydia pneumonia, helicobacter pylori, herpesviruses, and urinary tract infections with preeclampsia, a finding which could partially explain the association of infections with maternal stroke (52). In a population-based study in New York State, the diagnosis of an infection at the time of the admission was associated with a 3-fold increased risk of stroke in women with preeclampsia (35). Another study showed that women diagnosed with urinary tract infection or sepsis during their delivery admission had 2.5 times and 10.4 times the odds, respectively, of having a stroke during the admission, regardless of hypertensive status (53).

Physiological factors contributing to stroke risk during pregnancy and puerperium

Multiple physiologic factors inherent to pregnancy likely contribute to the increased risk of stroke during pregnancy and puerperium (Table 1). High levels of progesterone result in increased venous compliance and stasis, reaching a peak effect towards the end of pregnancy, coinciding with maximum direct compression of the fetus on the pelvic veins. Increased levels of estrogen result in higher production of procoagulant coagulation factors, including factors VII and X and prothrombin. Protein S levels decrease and there is an increase in activated protein C resistance in pregnancy. The placenta also produces plasminogen activator inhibitors, leading to a decrease in endogenous tissue plasminogen activator activity (54).

Intravascular volume expands during pregnancy from 30–50%, with a consequent 45% increase in cardiac output (55). These changes may have an impact on enlargement of brain arteriovenous malformations and cerebral aneurysms, although evidence is sparse and this is debated (19,56–59). Hemodynamic changes may exacerbate other conditions such as preexisting cardiomyopathies and valvular heart disease, contributing to increased risk of cardioembolic events, including stroke (60–62). Peripartum cardiomyopathy, a dilated cardiomyopathy highly associated with preeclampsia and affecting up to 1 in 3000 pregnancies, can lead to cardioembolic stroke (63). In a case series of 45 patients in Pakistan with peripartum cardiomyopathy, 9% developed embolic strokes within 6 months of diagnosis (64).

The inflammatory milieu of pregnancy and the puerperium may contribute to stroke risk. While pregnancy has classically been thought of as an immune-tolerant or even immunesuppressed state, the immunology of pregnancy is increasingly recognized to be far more complex than previously thought. Numerous theoretical models have been proposed to characterize the adaptation of the maternal immune system to the pregnant state, and have been comprehensively reviewed elsewhere (65). Pregnancy is associated with alterations in the innate immune response as well as a shift from Th1 to Th2 cell-mediated immunity. Increased expression of Th17, a highly inflammatory subtype of regulatory T cells, is also seen, particularly in the presence of infection (66,67). Similar inflammatory mechanisms have also been implicated in the pathogenesis of ischemic stroke independent of pregnancy (68).

Finally, pregnancy-related physiological changes in the cerebral vasculature may lead to vulnerability to ischemic or hemorrhagic stroke during pregnancy and postpartum. Animal studies have shown cerebral arteriolar remodeling and capillary proliferation in late stage pregnancy, with preserved autoregulatory function (69,70). Studies in preeclampsia rat models have demonstrated increased blood-brain barrier permeability, neuroinflammation, and cerebral autoregulatory dysfunction, possibly explaining the increased susceptibility to both hemorrhagic and ischemic stroke (71–73).

Specific mechanisms of stroke during pregnancy and postpartum

Specific causes of stroke in pregnancy and postpartum differ considerably from typical stroke mechanisms in other populations. Pathophysiological mechanisms contributing to maternal stroke are illustrated in Figure 1.

Figure 1. Mechanisms of maternal ischemic and hemorrhagic stroke.

Maternal strokes may occur through multiple pathophysiological mechanisms. Ischemic stroke mechanisms include cardioembolism (including paradoxical embolism), cervical artery dissection, RCVS, moyamoya stenoocclusive vasculopathy, amniotic fluid embolism, and arterial or venous thrombosis due to prothrombotic states such as APLS or sickle cell disease. Hemorrhagic stroke mechanisms include hypertensive hemorrhages due to preeclampsia, with or without HELLP-related coagulopathy; rupture of vascular lesions such as arteriovenous malformations, cerebral aneurysms, or moyamoya vessels; or hemorrhages related to venous sinus thrombosis. RCVS: reversible cerebral vasoconstriction syndrome. APLS: antiphospholipid syndrome. HELLP: hemolysis, elevated liver enzymes, and low platelets.

Ischemic stroke.

Ischemic stroke during pregnancy and postpartum can result either from arterial occlusion or venous infarction due to venous sinus thrombosis. Common arterial ischemic stroke mechanisms during pregnancy include cardioembolism (including paradoxical embolism from a PFO), cervical artery dissection, moyamoya-related stenoocclusive arterial disease, and arterial vasospasm due to reversible cerebral vasoconstriction syndrome (RCVS; see below). Direct atherosclerotic causes of arterial ischemic stroke are exceedingly rare in pregnant women (26).

Hemorrhagic stroke.

Hemorrhagic stroke includes non-traumatic ICH and SAH. Hemorrhagic infarction may also occur due to hemorrhagic transformation of ischemic strokes. Causes of hemorrhagic stroke in pregnancy include rupture of pre-existing vascular lesions such as brain arteriovenous malformations (AVMs), cerebral aneurysms, cerebral cavernous malformations, or fragile moyamoya collateral vessels. Hemorrhagic stroke may also result from pregnancy-related hypertension and coagulopathy, as can be seen in preeclampsia-eclampsia or the HELLP (hemolysis, elevated liver enzymes, and low platelets) syndrome, a particularly severe form of preeclampsia.

Posterior reversible encephalopathy syndrome (PRES) and reversible cerebral vasoconstriction syndrome (RCVS).

PRES and RCVS can cause both ischemic and hemorrhagic stroke in the peripartum and postpartum period. PRES is a syndrome of vasogenic brain edema thought to be precipitated by endothelial dysfunction and hypertension. Patients with PRES (both within and outside the context of pregnancy) typically present with encephalopathy and headaches; in severe cases, these symptoms are accompanied by seizures and visual alterations ranging from visual field defects to cortical blindness. Risk factors include untreated high blood pressure (particularly in the setting of preeclampsia), renal failure, history of autoimmune disorders, and exposure to cytotoxic medications. Several hypotheses point to increased endothelial permeability as a main feature of this syndrome. This may be due to direct action of inflammatory cytokines, or due to a rapid increase in blood pressure which overcomes cerebral autoregulatory capacity (74).

RCVS is a condition of alteration of intracerebral vascular tone resulting in diffuse, but transient cerebral vasospasm. RCVS was first described in postpartum women by Call and Fleming and in the setting of pregnancy is often termed “postpartum angiopathy” (75). RCVS is also seen outside of pregnancy, often associated with the use of vasoactive drugs including sympathomimetics or serotonergic medications, or neuroendocrine tumors. Approximately 7–9% of RCVS cases occur in postpartum women (76,77). Patients typically present with recurrent episodes of “thunderclap” headaches, reaching maximal intensity in less than 60 seconds. Focal neurological deficits may be present in severe cases. Imaging studies usually show multifocal areas of stenosis limited to the intracranial vasculature, which resolve within 12 weeks of onset (78).

There is considerable overlap between the syndromes of PRES, RCVS, and preeclampsia-eclampsia. In one series, 98% of patients with eclampsia had radiographic evidence of PRES on magnetic resonance imaging, and some authors consider eclampsia and PRES indistinguishable (79,80). Others have speculated that pregnancy-related RCVS and PRES may have a different pathophysiology, possibly facilitated by neurovascular unit dysfunction particularly during the third trimester of gestation and postpartum (27). While the pathophysiological disturbances leading to PRES and RCVS are temporary and reversible, they may lead to permanent damage from ischemic and/or hemorrhagic stroke (Figure 2).

Figure 2. Postpartum stroke.

1. A 23 year old woman presented on postpartum day 5 with recurrent “thunderclap” headaches and focal neurological deficits. CT angiography demonstrated multifocal stenoses in the (A) anterior cerebral artery and (B) middle cerebral artery, consistent with reversible cerebral vasoconstriction syndrome.

2. A 28 year old woman with preeclampsia presented on postpartum day 2 with severe headache and altered mental status. CT head without contrast showed a large right frontal intracerebral hemorrhage with intraventricular extension. CT angiography showed no underlying vascular abnormality.

Cerebral venous thrombosis.

CVT has an annual incidence of 1.57 per 100,000 person-years, and is more common in women, with a 3:1 female-to-male ratio (81,82). Risk factors include hormonal therapy, thrombophilia, dehydration, head trauma, infections, hypotension and the postpartum state. In a systematic review and pooled analysis of pregnancy-related CVT, the most frequently involved sinuses are the superior sagittal sinus (67%) and the transverse sinuses (64%) (83). Headache is the most common presenting symptom. Seizures, focal neurological deficits, and lethargy are usually seen later in the course, if left untreated. The diagnosis of CVT may be missed during the puerperium, due to being mistaken for migraine or post-dural puncture headache. CVT-related headache can often be differentiated from other headache types due to its positional quality, with headache more severe in the supine position due to increased intracranial pressure, and its progressive course if untreated.

CVT leads to elevated venous pressure and decreased cerebrospinal fluid absorption, both of which can result in venous infarction, ICH and SAH. Elevated intracranial pressure may rapidly develop, leading to focal deficits, seizure or coma. The treatment for CVT is anticoagulation, even in the presence of intracranial hemorrhage.

Acute Management of Maternal Stroke

Maternal stroke is a medical emergency requiring rapid, multidisciplinary management. Few high-quality data are available to inform management. The Canadian Heart and Stroke Foundation recently published guidelines for management of acute stroke in pregnancy based on the best available evidence and expert opinion (23). Neurologists, emergency physicians, obstetricians, obstetric anesthesiologists, and interventional neuroradiologists and/or neurosurgeons may all be involved in the rapid assessment and decision-making required in this situation.

Diagnostic imaging

Non-contrast studies, including computed tomography (CT) and magnetic resonance imaging (MRI), are not contraindicated during pregnancy. While CT uses ionizing radiation, the benefit of rapid diagnosis during an acute stroke evaluation generally outweighs any theoretical fetal risk, making it the test of choice in emergent evaluation of acute stroke. The estimated fetal radiation exposure for a single maternal CT head is 0.001 to 0.01 mGy, far below the safety threshold for teratogenesis or developmental disabilities, estimated at 50mGy. MRI brain is considered safe during pregnancy, with no evidence of fetal harm (84).

Contrasted studies.

Contrast agents should be avoided during pregnancy, if possible. However, iodinated contrast, classified as an FDA class B drug during pregnancy, is preferred in emergent situations when non- contrasted studies cannot provide adequate information to ensure an accurate diagnosis and treatment; there is no evidence of fetal harm from iodinated contrast in humans. Breastfeeding can also be continued after the administration of iodinated contrast. Gadolinium is a class C drug; its use is not recommended during gestation as it crosses the placenta and has been associated with teratogenesis in animal studies (84). Additional techniques potentially of use in this population include MRI with arterial spin labeling, which avoid the use of ionizing radiation or contrast, instead using the patient’s own blood as the “contrast” agent and providing a map of the collateral circulation, areas of occlusion and perfusion maps (85,86).

Management of acute ischemic stroke in pregnancy and postpartum

Pregnant patients were excluded from thrombolytics and thrombectomy trials and available evidence is limited to a few retrospective studies and case series. Due to fear of possible fetal complications, pregnant and postpartum women may be offered fewer options in the acute setting. However, the evidence supports use of all available treatment options in most situations.

Intravenous thrombolysis.

Tissue plasminogen activator (alteplase) is a thrombolytic classified as category C by the FDA. Alteplase does not cross the placenta after intravenous infusion due to its large molecular weight of 59,000 daltons (87). The half -life of the medication is 4–5 minutes with less than 10% of the medication present in the circulation after 20 minutes (88). Animal studies have reported possible teratogenicity; however, the events were seen only with dosages of alteplase equivalent to at least three times the therapeutic dose in humans (87). Favorable outcomes have been reported after thrombolysis for ischemic stroke in pregnant women (15,89–91). The American Heart Association/American Stroke Association recommends that pregnant women be considered for thrombolytic therapy when the perceived benefits outweigh the risks, but found insufficient evidence to make a recommendation regarding thrombolysis in the postpartum period (92).

Mechanical thrombectomy.

While pregnant women have been excluded from all randomized thrombectomy trials, multiple cases have been reported of pregnant women who underwent successful thrombectomy for acute ischemic stroke and subsequently had an uneventful recovery with good functional recovery and no subsequent delivery or infant complications (91,93,94). A registry-based study showed that women who had thrombolysis and/or thrombectomy for acute ischemic stroke during pregnancy had similar outcomes compared to non-pregnant patients, despite having more severe strokes at baseline (15). The Canadian guidelines recommend considering endovascular thrombectomy as the benefits of the procedure almost always outweigh the risk, considering the severe fetal and maternal morbidity associated with large vessel occlusion strokes when left untreated (23).

Management of hemorrhagic stroke in pregnancy

Management of hemorrhagic stroke in pregnancy is similar to management in a non-pregnant patient. Initial management after the diagnosis of intracranial hemorrhage of any type is aimed at prevention of re-bleeding and hematoma expansion. Coagulopathies should be corrected and the source of the hemorrhage should be promptly identified with CT angiography and/or diagnostic cerebral angiogram. Immediate blood pressure control following the diagnosis of ICH is a priority. Close obstetrical observation including fetal monitoring is needed to monitor for signs of placental hypoperfusion. Labetalol, methyldopa, nicardipine, and long acting nifedipine may all be used safely in pregnant women (23).

Arteriovenous malformations.

Neurosurgical and/or endovascular treatment should be initiated as soon as AVM is identified as a cause of intracranial hemorrhage, due to the risk of rapid clinical deterioration. Cerebral angiogram and embolization with appropriate shielding are considered safe for the fetus (18). Strict blood pressure control during anesthesia is recommended, to avoid expansion of the hematoma while avoiding placental hypoperfusion (23).

Management of unruptured, asymptomatic AVMs is more controversial; the only randomized trial of intervention versus medical management for unruptured AVMs was stopped due to evidence of harm in the intervention arm, and long term results continue to show better functional outcomes in the medically managed group (95,96). However, the unique risks related to pregnancy were not addressed in that trial. Conservative management is usually recommended in the absence of intracranial hemorrhage, although some experts recommend preemptive treatment given the special risks of pregnancy (97). In patients with high-risk features including deep location and drainage, arteriovenous fistula, or concomitant aneurysm, the timing of any planned intervention should be decided on an individualized basis (18).

Cerebral aneurysms.

Recommendations for ruptured intracranial aneurysms also favor rapid intervention. Endovascular or open procedures to secure the aneurysm are recommended after the culprit lesion is identified. Delaying interventions or diagnostic testing due to fear of possible radiation or contrast effects has proven to be detrimental for mother and the fetus (56,98). To date, most evidence has shown no increased rupture of cerebral aneurysms during pregnancy, and conservative treatment of unruptured aneurysms is recommended in most cases (97).

Hypertensive and PRES/RCVS related hemorrhages.

Management of hypertensive hemorrhages should focus on blood pressure control (39). Preeclampsia-eclampsia should be treated per current obstetric guidelines, including initiation of magnesium sulfate therapy and rapid delivery of the fetus (99). Emergent clot evacuation or placement of an external ventricular drain may be needed in some cases. For PRES and RCVS related hemorrhages, steroids should be avoided as they may worsen outcomes (100,101).

Delivery considerations after stroke.

Delivery after pregnancy-related stroke should be guided by obstetric indications. Cesarean delivery has not been shown to improve outcomes and is associated with an increased risk of peripartum stroke of all types (33). A multidisciplinary approach to maternal-fetal care after stroke is recommended, including maternal-fetal medicine specialists, neurologists, and obstetric anesthesiologists.

Recurrence of maternal stroke in future pregnancies

Few data are available regarding the recurrence rate of maternal stroke in future pregnancies. One study found that women with a history of prior ischemic stroke (not pregnancy-associated) had an increased risk of recurrent stroke in the postpartum period, but not during pregnancy itself (102). However, the absolute risk of recurrent ischemic stroke was less than 2%. Two case series of 23 and 24 women, respectively, both found a 0% recurrence rate for ischemic stroke in future pregnancies (103,104). A study from a large international CVT registry found a recurrence rate for CVT of 12 per 1000 deliveries (95%CI, 2–66) (105). While recurrent hemorrhagic stroke due to postpartum RCVS has been reported, no large studies have been published (106). Guidelines recommend continuing tailored secondary stroke prevention and treating risk factors such as hypertension aggressively during pregnancy in women with a history of stroke (107).

Conclusions and future directions

Maternal stroke is increasingly recognized as an important cause of maternal morbidity and mortality, both of which are rising in the US. However, many gaps remain in our understanding of the factors that may precipitate maternal stroke. For example, while various risk factors have been identified for maternal stroke, there are no prediction tools to help identify which women might be at highest risk for postpartum stroke and require closer monitoring. The risk of recurrent maternal stroke, especially hemorrhagic stroke, has not been adequately quantified, making it difficult for clinicians to counsel patients. The complex pathophysiology of preeclampsia and other HDP, and its effects on the cerebral vasculature, requires further targeted study with animal models. Similarly, the pathophysiological link between migraine and maternal stroke is not well understood. The mechanisms by which infection may increase the risk of maternal stroke, and the role of the maternal immune system, remain incompletely characterized. Genetic and genomic factors contributing to maternal stroke have not been investigated.

Fortunately, there is mounting interest in the clinical and research community in understanding the pathophysiology of maternal stroke. The number of publications focusing on pregnancy and stroke has proliferated steadily over the past 20 years. The issue of maternal morbidity and mortality has garnered national attention and prompted the formation of publicly funded programs such as New York State’s Safe Motherhood Initiative (108). The time is ripe for an increased focus on the prevention, recognition, and optimal treatment of maternal stroke as a critical aspect of improving maternal outcomes.