The etiology of eclamptic convulsions is unknown, and there are many unanswered questions regarding the pathogenesis of the cerebral manifestations. Several theories and etiologic mechanisms have been implicated as possible etiologic factors, but none of these have been conclusively proven.

Cerebral autoregulation is a mechanism for maintenance of constant cerebral blood flow during changes in blood pressure; this is hypothesized to be altered in eclampsia. Cerebral blood flow normally remains relatively constant when cerebral perfusion pressure ranges between 60 and 120 mmHg [9]. In this normal range, the elevations in blood pressure result in vasoconstriction of the cerebral vessels, whereas vasodilation occurs as BP decreases.

Studies with Doppler ultrasound have outlined a picture of the cerebral hemodynamics in both normal pregnancy and preeclampsia. In normal pregnancy, the systolic velocity and resistance index in the middle cerebral artery both decrease approximately 20 % over gestation [9], whereas the cerebral perfusion pressure (CPP) increases by 50 % from early pregnancy to term (Figs. 7.1, 7.2, 7.3, and 7.4).

The Doppler-derived data have also been confirmed by magnetic resonance imaging studies that show that the middle and posterior cerebral artery diameters remain static during normal late pregnancy, whereas the flow (which in this stance is proportional to the velocity) decreased by approximately 20 % [123]. Cerebral autoregulation is very efficient during pregnancy, and despite a significant increase in perfusion pressure, the cerebral blood flow changes by a much smaller percentage. A small decrease in cerebral resistance is seen in normal pregnancy, as blood pressure increases within the normal range, and this is believed to the result of prostacyclin release as the vessel walls are distended. However, as pressure increases out of the normal range (in preeclamptic women without headache), a physiologic increase in cerebral resistance occurs to limit perfusion and should not be regarded as pathologic change [11, 12] (Fig. 7.5).

Once cerebral perfusion pressure exceeds 130–150 mmHg, the autoregulatory mechanism fails [107]. The normal compensatory vasoconstriction may fail in extreme hypertension resulting in cerebral overperfusion, which may or may not be accompanied by vasospasm and ischemia when vascular integrity is breached [9, 13, 78]. Belfort et al. compared cerebral perfusion in 72 women with preeclampsia without severe features and 120 women with preeclampsia with severe features [13] (Fig. 7.6). A significant proportion of severe preeclamptics have high perfusion pressure (52 %) versus only a minority of preeclamptic patients without severe features. Overall they showed that in preeclampsia with severe features, the resistance is abnormally high, whereas it is within the normal range in women with mild preeclampsia [13].

As a result of overperfusion, segments of the cerebral vessel are believed to become dilated and increasingly permeable with exudation of plasma, leading to focal cerebral edema, compression of brain tissue and blood vessels, and ultimately decreased cerebral blood flow [107].

Hypertensive encephalopathy and cerebral overperfusion are now believed to be the more likely model for the most cases of eclampsia not associated with hemorrhage, as opposed to cerebral ischemia and vasospasm [9, 106]. Hypertensive encephalopathy is an acute clinical condition resulting from abrupt severe hypertension and subsequent significant increases in intracranial pressure [122, 123]. On the basis of cerebral imaging, hypertensive encephalopathy and eclampsia share many clinical, radiologic, and pathologic features. In patients with hypertensive encephalopathy and some patients with eclampsia, there is failure of normal cerebral blood flow autoregulation [33, 34, 92]. There are two theories proposed for these cerebral abnormalities: forced dilation and vasospasm [34]. The forced dilation theory suggests that the vasogenic edema frequently seen in eclampsia is caused by a loss of cerebrovascular autoregulation [11, 12, 111]. Normal physiologic cerebral vasoconstriction initially controls downstream pressure and volume flow with increasing blood pressure. When the upper limit of autoregulation is reached, forced vasodilation starts to occur as the capacity of the local artery is overwhelmed (either by the duration or extent of the hypertensive stimulus), allowing for local overperfusion with subsequent interstitial or vasogenic edema [11, 12, 34, 111]. The vasospasm theory proposed that cerebral overregulation occurs in response to acute severe hypertension with resultant cerebral underperfusion, ischemia, cytotoxic edema, and infarction [34, 112, 122]. Recently Van Veen et al. conducted a study comparing dynamic cerebral autoregulation in 20 patients with preeclampsia versus 20 healthy pregnancy women [113]. They showed that impaired dynamic cerebral autoregulation does not correlate with blood pressure (corroborating the same finding for static autoregulation demonstrated by Belfort et al. [9, 11, 12]), which might explain why cerebral complications such as eclampsia can occur without sudden or excessive blood pressure.

Hinchey et al., in 1996, linked eclampsia to a condition that they called reversible posterior leukoencephalopathy syndrome [46]. This syndrome comprised a variety of symptoms and signs including headache, visual disturbances, altered mental status, and seizures, as well as radiologic features of cerebral edema, which was mainly seen in the occipital lobes and posteriorly in the brain. This concept has persisted, but the syndrome itself was renamed as posterior reversible encephalopathy syndrome (PRES) [44] (Fig. 7.7).

The reason PRES seems to mainly affect the parieto-occipital lobes is not known at present [42, 46, 53, 92]. It is possible that this may be related to decreased sympathetic innervation in the vertebrobasilar arteries (as compared to the internal carotid arteries) [56], leading to overwhelming of the autoregulatory capacity during acute hypertension at a lower pressure than in those areas that have more dense sympathetic innervation [25, 56].

The long-term consequences of eclampsia and preterm preeclampsia (gestational age <37 weeks) have been studied by Aukes et al. who have shown that remote preterm preeclampsia is associated with an increased prevalence of cerebral white matter lesions on MRI when compared to control patients (who had normotensive gestation or term preeclampsia) [6, 7].

It is difficult to make a case that these white matter lesions were caused by PRES, however, since the lesions are mainly found in the frontal part of the brain (not only in the posterior regions) and also occur in patients who did not have seizures. It is more likely that there is an underlying predisposition for preeclamptic women to develop cerebrovascular disease in later life [119].

In summary, many women with eclampsia will have evidence of vasogenic edema on brain imaging usually in the occipital and parietal lobes. While not confirmatory, this provides good circumstantial evidence that hypertensive encephalopathy plays a central role in the pathogenesis of eclamptic convulsions. The long-term consequences of eclampsia (and severe preeclampsia) are unclear, but there are data to suggest last effects.

Convulsions in association with hypertension (with or without proteinuria) in a currently, or recently, pregnant woman suggest the diagnosis of eclampsia [107]. The hypertension may not necessarily be in the severe range, and a high level of suspicion is required so as not to overlook this diagnosis. In patients, who develop eclampsia, there is a wide spectrum of signs reported, ranging from no preceding signs to mild to severe hypertension, minimal to severe proteinuria, and absent to generalized edema [104, 107]. Some degree of hypertension is almost always present but may be absent in 16 % of the cases [74]. Hypertension may be severe (systolic ≥160 mmHg and/or diastolic ≥110 mmHg) in 20–54 % of cases [37, 74] or mild (systolic blood pressure between 140 and 160 mmHg and diastolic between 90 and 110 mmHg) in 30–60 % of cases [37, 74]. Severe hypertension is more common in patients who develop eclampsia in the antepartum period (58 %) and especially in cases who develop eclampsia at 32 weeks or earlier (71 %) [74].

Proteinuria (protein to creatinine ratio ≥0.3 or ≥300 mg per 24 h urine collection or dipstick at least +1 only if other quantitative methods are not available) may also be associated with eclampsia [74]. Mattar and Sibai showed in a series of 399 patients with eclampsia that severe proteinuria (equal or greater than 3+ on dipstick) only occurred in 48 % of the cases and that proteinuria was absent in 14 % [74]. Abnormal weight gain (with or without clinical edema) in excess of 2 lbs per week in the third trimester might be the only sign preceding eclampsia. In the same study, edema was absent in 26 % of the 399 eclamptic patients [74].

Most women have premonitory symptoms in the hours before the initial seizure. These include persistent occipital or frontal headaches, blurred vision, photophobia, epigastric or right upper quadrant pain, and altered mental status, and 59–75 % of women who develop eclampsia have at least one of these symptoms (Table 7.2). Patients report headache as their major preceding complaint in 50–75 % of cases, whereas visual changes have been reported in 19–32 % of the patients [23, 37, 50]. In a systematic review that included 59 studies involving more than 21,000 women with eclampsia from 26 countries, the most common antecedent signs and symptoms were hypertension (75 %), headache (66 %), visual disturbances (scotomata, loss of vision [cortical blindness], blurred vision, diplopia, visual field defect, photophobia) (27 %), and right upper quadrant or epigastric pain (25 %) [14]. In 25 % of cases, the patients were asymptomatic prior to the seizure(s) [14].

Eclampsia is generally manifested by a generalized tonic-clonic seizure or by coma. At the onset, there is usually an abrupt loss of consciousness, often associated with a scream or shriek, and muscle stiffening in the arms, legs, chest, and back [94]. The patient may appear cyanotic during this tonic phase, which may last from a few seconds to approximately a minute. Following the hypertonicity, the patient’s muscles will usually begin to jerk or twitch for an additional 1–2 min. During this clonic phase, the patient may bite her tongue and express frothy and bloody sputum. The postictal phase begins once the twitching movements end. This is usually followed by a deep sleep, deep breathing, and gradual return to sentience. On awakening the patient will often complain of a headache. Most patients begin to recover responsiveness within 10–20 min after the generalized convulsion. Focal neurologic deficits are generally absent although there may be memory deficits. On examination there may be increased deep tendon reflexes, visual perception deficits, altered mental status, and cranial nerve deficits [94].

Eclamptic convulsion can be antepartum, intrapartum, or postpartum. The frequency of antepartum eclamptic convulsion has been reported to be 38–53 % [23, 37, 50, 74]. Mattar and Sibai reported that most cases of eclampsia develop at or beyond 28 weeks (91 %). The same study showed that 7.5 % of patients who develop eclampsia develop it between 21 and 27 weeks of gestation and that in 1.5 % eclampsia occurs at 20 weeks of gestation or earlier [74].

If eclampsia develops before 20 weeks of gestation, the coexistence of molar pregnancy needs to be ruled out [81, 102]. Although rare, several case reports have described eclampsia during the first trimester without the coexistence of molar pregnancy [74, 81], and for this reason, eclampsia should be ruled out at any gestation in pregnant women who have had a seizure [102]. Women with early eclampsia may be misdiagnosed with a seizure disorder, hypertensive encephalopathy, or thrombotic thrombocytopenia purpura. Women in whom a convulsion(s) is associated with hypertension and/or proteinuria during the first half of pregnancy should thus be considered eclampsia unless proven otherwise [104]. All women with early gestation eclampsia (first half of pregnancy) should have ultrasound examination of the uterus to rule out a molar pregnancy. Also extensive medical evaluation and neurologic examination should be performed for these patients to rule out other etiologies such as meningitis, cerebral abscess, encephalitis, cerebral hemorrhage or thrombosis, cerebral vasculitis, thrombotic thrombocytopenia purpura (TTP), brain tumor, and metabolic disease; it is always wise to rule out chemical and drug (legal and illicit) exposures [104, 107].

The incidence of postpartum eclampsia ranges from 11 to 44 %, and most of the postpartum eclampsia occurs during the first 48 h following delivery [23, 37, 50, 74, 102]. However, eclampsia can and does develop beyond 48 h and has been reported as late as 23 days postpartum [23, 50, 74]. Late postpartum eclampsia is defined as eclampsia that occurs beyond 48 h but within 4 weeks of delivery [62, 102]. Approximately 56 % of these women will demonstrate signs and symptoms of preeclampsia during labor or immediately postpartum, whereas others will show these clinical findings for the first time more than 48 h after delivery (44 %) [62]. Late postpartum eclampsia can develop despite the use of prophylactic intra- and postpartum (at least 24 h) magnesium sulfate in previously diagnosed women with preeclampsia [23, 62]. Therefore, women with convulsion(s) associated with hypertension and/or proteinuria and/or with headaches or visual disturbances 48 or more hours after delivery should be considered to have eclampsia unless otherwise proven and should be treated as for eclampsia [23, 62, 102]. In these cases, an extensive neurological evaluation including CNS examination, cerebrovascular testing, and brain imaging (MRI and/or CT depending on the circumstances) and routine laboratory tests (CBC and platelets, liver function, renal function and electrolytes, and coagulogram) are usually instituted at a minimum, with further more sophisticated testing (lumbar puncture, EEG, and angiography) used as dictated by initial findings [23, 62, 102].

Several neurodiagnostic tests, such as electroencephalography (EEG), computerized tomography (CT), magnetic resonance imaging (MRI), diffusion-weighted magnetic resonance imaging (DWI), cerebral Doppler velocimetry, and cerebral angiography (both traditional and MRI angiography), have been studied in women with eclampsia.

There is limited information on EEG in eclampsia. In general, despite the fact that the EEG is almost always acutely abnormal in patients with eclampsia, none of the identified patterns are pathognomonic for eclampsia [107]. Review of the EEG literature has shown that postictal EEG abnormalities are common in eclamptic women and the EEG almost always becomes normal with prolonged postpartum follow-up [18]. The reliability of these studies has been questioned given methodological issues, and the fact that all but one were published between 1955 and 1984. There are no more recent EEG studies available in which more modern equipment and practices have been used.

Lumbar puncture is not helpful in the diagnosis and management of patients with eclampsia and may be dangerous if the patient has acutely elevated intracranial pressure. For this reason, such a test should only be ordered when the differential diagnosis absolutely requires it and the benefit outweighs the risk [107].

CT and MRI studies performed following seizure activity in preeclamptic patients usually reveal the presence of edema and infarction within the subcortical white matter and adjacent gray matter (mostly in the parietal and occipital lobes). Other CT and MRI findings have been summarized in Table 7.3.

In uncomplicated eclampsia (i.e., no cerebral hemorrhage, hydrocephalus, or congenital anomaly), cerebral imaging findings are similar to those found in patients with hypertensive encephalopathy. The classic findings are referred to as posterior reversible encephalopathy syndrome (PRES) [53]. In a small series of eclamptic patients studied by Brewer and colleagues, 46 of 47 patients (97.9 %) revealed PRES on CT or MRI with or without contrast [17].

Within the past two decades, magnetic resonance diffusion-weighted imaging (DWI) and apparent diffusion coefficient (ADC) mapping have become more commonly used and reported, and this facilitates the discrimination between vasogenic and cytotoxic forms of cerebral edema [24]. DWI takes advantage of strong diffusion gradients that detect changes in water molecule distribution in cerebral tissue. In the presence of infarction, cytotoxic edema is caused by sodium pump failure, and the resultant reduction in proton diffusion elicits hyperintense (“bright”) signal on DWI (Fig. 7.8). Conversely, vasogenic edema is characterized by increased extracellular fluid with enhanced water diffusion, and this may be seen as normal or decreased signal brightness on DWI. While it is generally a reliable way of distinguishing vasogenic from cytotoxic edema, occasionally a hyperintense signal may be seen in DWI in patients who do not have cytotoxic edema, and this has been dubbed “T2 shine-through” (Figs. 7.9 and 7.10). Thus, specialist in radiology expertise may be required to determine whether DWI hyperintensity is due to restricted diffusion or to T2 shine-through, prior to instituting specific therapy or counseling regarding a prognosis.

This issue is usually resolved by estimation of the underlying ADC in the region of interest since ADC map is independent of T2 effects and can be used to determine whether diffusion is restricted or free in the area of interest. A decreased ADC map that corresponds to hyperintense areas on the DWI confirms restricted diffusion, while an elevated ADC results from water molecules with increased diffusional motion and thus represents vasogenic edema.

In two small series [61, 123], the frequency of vasogenic edema and cytotoxic edema in eclamptic patients was estimated. Cerebral edema (mostly vasogenic) was present in up to 93–100 % of these women. Concurrent foci of infarction and cytotoxic edema, as evidenced by reduced apparent diffusion coefficient (restricted diffusion), were present in six of 27 patients studied by Zeeman and colleagues [123] and in three of 17 eclamptic and preeclamptic women studied by Loureiro and associates [61]. In addition five of six women reported by Zeeman and colleagues [123] and four of 17 women reported by Loureira et al. [61] had persistent abnormalities on repeat MRI testing 6–8 weeks later, suggesting that these lesions might not be reversible.

Uncomplicated eclampsia (full recovery) is usually a clinical diagnosis and does not require cerebral imaging diagnosis and management. However, cerebral imaging is indicated for patients with focal neurologic deficits, prolonged coma, fever, suspicion of TTP, or other imitators of preeclampsia and those who develop seizures with a therapeutic MgSO₄ level or who are unresponsive to MgSO₄ therapy [107]. We also recommend that all postpartum eclampsia be investigated with cerebral imaging.

In these patients, hemorrhage and other serious abnormalities requiring specific pharmacologic therapy or surgery must be excluded. Cerebral imaging also might be helpful in cases of atypical eclampsia including normotensive and/or non-proteinuric eclampsia, onset of eclampsia before 20 weeks of gestation (after excluding molar pregnancy) or after delivery, and in women who may have known antiphospholipid syndrome or autoimmune disease [107]. Advances in MRI and magnetic resonance angiography, as well as in cerebral vascular Doppler velocimetry, may aid our understanding regarding the pathogenesis and improving long-term outcome of this condition [106].

It is important to distinguish PRES (an overperfusion syndrome with forced cerebral vasodilatation) from reversible cerebral vasoconstriction syndrome (RCVS), which is different in etiology (vasospasm) [39]. RCVS is characterized by recurrent thunderclap headaches, seizures, strokes, and nonaneurysmal subarachnoid hemorrhage [39]. RCVS seems to be associated with constriction and/or dilation of large or medium arteries, while PRES does at the level of distal arterioles and capillary. An overlap between these two syndromes represents a continuum in them. Postpartum cerebral angiopathy is another ill-characterized RCVS, usually occurring within 30-day duration of uncomplicated pregnancy and delivery. The diagnosis is confirmed by angiography (Fig. 7.11) [39].

As discussed above, a differential diagnosis (Table 7.4) should be considered. The diagnosis and management of these diagnoses are beyond the scope of this chapter.