Cerebral infarcts, edema, hypoperfusion, and vasospasm in preeclampsia and eclampsia

Eclampsia is a serious pregnancy complication and is associated with cerebral edema and infarctions. However, the underlying pathophysiology of eclampsia remains poorly explored.

This study aimed to assess the pathophysiology of eclampsia using specialized magnetic resonance imaging to measure diffusion, perfusion, and vasospasm.

This was a cross-sectional study recruiting consecutive pregnant women between April 2018 and November 2021 at Tygerberg Hospital, Cape Town, South Africa. Women with eclampsia, preeclampsia, and normotensive pregnancies who underwent magnetic resonance imaging after birth were recruited. The main outcome measures were cerebral infarcts, edema, and perfusion using intravoxel incoherent motion imaging and vasospasm using magnetic resonance imaging angiography. The imaging protocol was established before inclusion.

Here, 49 women with eclampsia, 20 women with preeclampsia, and 10 normotensive women were included. Cerebral infarcts were identified in 34% of women with eclampsia and 5% of women with preeclampsia (risk difference, 0.29; 95% confidence interval, 0.06-0.52; P=.012). However, no cerebral infarct was identified in normotensive controls. Women with eclampsia were more likely to have vasogenic cerebral edema than women with preeclampsia (80% vs 20%, respectively; risk difference, 0.60; 95% confidence interval, 0.34-0.85; P<.001) and normotensive women (risk difference, 0.80; 95% confidence interval, 0.47-1.00; P<.001). Diffusion was increased in women with eclampsia in the parieto-occipital white matter (mean difference, 0.02 × 10^-3 mm²/s; 95% confidence interval, 0.00-0.05; P=.045) and caudate nucleus (mean difference, 0.02 × 10^-3 mm²/s; 95% confidence interval, 0.00-0.04; P=.033) compared with women with preeclampsia. In addition, diffusion was increased in women with eclampsia in the frontal white matter (mean difference, 0.07 × 10^-3 mm²/s; 95% confidence interval, 0.02-0.12; P=.012), parieto-occipital white matter (mean difference, 0.05 × 10^-3 mm²/s; 95% confidence interval, 0.02-0.07; P=.03), and caudate nucleus (mean difference, 0.04 × 10^-3 mm²/s; 95% confidence interval, 0.00-0.07; P=.028) compared with normotensive women. Perfusion was decreased in edematous regions. Hypoperfusion was present in the caudate nucleus in eclampsia (mean difference, -0.17 × 10^-3 mm²/s; 95% confidence interval, -0.27 to -0.06; P=.003) compared with preeclampsia. There was no sign of hyperperfusion. Vasospasm was present in 18% of women with eclampsia and 6% of women with preeclampsia. However, no vasospasm was present in the controls.

Eclampsia was associated with cerebral infarcts, vasogenic cerebral edema, vasospasm, and decreased perfusion, which are not usually evident on standard clinical imaging. This finding may explain why some patients have cerebral symptoms and signs despite having normal conventional imaging.