ABSTRACT
Objective
Pre‐eclampsia is a hypertensive complication of pregnancy that is associated with an increased risk of long‐term cardiovascular and cerebrovascular disorders. Although the underlying mechanism of persistent susceptibility to cerebral complications after pre‐eclampsia remains largely unclear, impaired blood–brain barrier (BBB) integrity has been suggested to precede several cerebrovascular diseases. In this study, we aimed to investigate the integrity of the BBB years after pre‐eclampsia.
Methods
This was an observational study of premenopausal formerly pre‐eclamptic women and controls with a history of normotensive pregnancy who underwent cerebral magnetic resonance imaging (MRI) at ultra‐high field (7 Tesla) to assess the integrity of the BBB. Permeability of the BBB was determined by assessing leakage rate and fractional leakage volume of the contrast agent gadobutrol using dynamic contrast‐enhanced MRI. BBB leakage measures were determined for the whole brain and lobar white and gray matter. Multivariable analyses were performed, and odds ratios were calculated to compare women with and those without a history of pre‐eclampsia, adjusting for potential confounding effects of age, hypertension status at MRI and Fazekas score.
Results
Twenty‐two formerly pre‐eclamptic women (mean age, 37.8 ± 5.4 years) and 13 control women with a history of normotensive pregnancy (mean age, 40.8 ± 5.5 years) were included in the study. The time since the index pregnancy was 6.6 ± 3.2 years in the pre‐eclamptic group and 9.0 ± 3.7 years in controls. The leakage rate and fractional leakage volume were significantly higher in formerly pre‐eclamptic women than in controls in the global white (P = 0.001) and gray (P = 0.02) matter. Regionally, the frontal (P = 0.04) and parietal (P = 0.009) cortical gray matter, and the frontal (P = 0.001), temporal (P < 0.05) and occipital (P = 0.007) white matter showed higher leakage rates in formerly pre‐eclamptic women. The odds of a high leakage rate after pre‐eclampsia were generally higher in white‐matter regions than in gray‐matter regions.
Conclusion
This observational study demonstrates global impairment of the BBB years after a pre‐eclamptic pregnancy, which could be an early marker of long‐term cerebrovascular disorders. © 2022 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.
Keywords: 7 Tesla, blood–brain barrier, cerebrovascular disorder, dynamic contrast‐enhanced MRI , pre‐eclampsia
CONTRIBUTION —
What are the novel findings of this work?
Dynamic contrast‐enhanced MRI at 7 Tesla revealed subtle blood–brain barrier (BBB) leakage several years after a pre‐eclamptic pregnancy. The BBB leakage rate and fractional leakage volume were globally higher in formerly pre‐eclamptic women than in control women, indicating diffusively spread endothelial dysfunction as a pathophysiological mechanism.
What are the clinical implications of this work?
Pre‐eclampsia is associated with an increased cerebrovascular risk. BBB leakage often precedes cerebrovascular disorders. This study demonstrates global BBB disruption years after pre‐eclampsia. The BBB could be a novel target for treatment and prevention strategies aimed at reducing the cerebrovascular risk after pre‐eclampsia.
INTRODUCTION
Pre‐eclampsia is a hypertensive complication of pregnancy that affects multiple organ systems 1 . A history of pre‐eclampsia raises the long‐term risk of cerebrovascular disorders 2 , 3 . Impaired cerebrovascular function during a pre‐eclamptic pregnancy is thought to originate from decreased cerebrovascular resistance as a consequence of the loss of cerebral blood flow autoregulation, which is accompanied by hyperperfusion, blood–brain barrier (BBB) disruption and vasogenic edema 4 , 5 , 6 , 7 , 8 , 9 . An increase in BBB permeability has been demonstrated in animal models exposed to blood plasma from women with pre‐eclampsia 10 , 11 , 12 , 13 , 14 . However, studies assessing BBB permeability in pre‐eclamptic women without overt neurological disturbances are lacking, and potential long‐term effects of pre‐eclampsia on the BBB have not been assessed.
The BBB plays an important role in maintaining the homeostasis of the brain environment, which is essential for brain health 15 . BBB dysfunction may create an unfavorable microenvironment with eventual neuronal dysfunction and has been associated with a decrease in cognitive functioning 15 , 16 . Given that formerly pre‐eclamptic women report subjective cognitive and psychological problems years after a pregnancy complicated by pre‐eclampsia 17 , 18 , the long‐term condition of the BBB may provide new insights into the underlying pathophysiological mechanism of persistent susceptibility to cerebral disorders in this population.
Localized BBB dysfunction can be investigated using dynamic contrast‐enhanced magnetic resonance imaging (DCE‐MRI) 19 by measuring the leakage of paramagnetic contrast agent into the brain tissue 20 , 21 . While some pathologies, such as high‐grade tumors 22 and multiple sclerosis 23 , show relatively high contrast agent leakage into a lesion, other disorders, such as cerebral small vessel disease (cSVD), have shown subtler contrast agent leakage 24 . Considering the cognitive 18 and psychological 17 problems and white‐matter (WM) abnormalities 25 , 26 , 27 observed in patients following pre‐eclampsia, which are similar to the symptoms observed in early cSVD, subtle contrast agent leakage from the cerebral microvasculature into the brain tissue may also be present in formerly pre‐eclamptic women. Ultra‐high field (7‐Tesla) MRI, instead of clinically used 1.5‐ or 3‐Tesla MRI, can be employed to aid in the detection of subtle BBB leakage 28 , 29 . In this study, we used a dedicated 7‐Tesla DCE‐MRI protocol to investigate differences in BBB permeability between formerly pre‐eclamptic women and controls with a history of normotensive pregnancy.
METHODS
Participants and procedure
This was an observational study of premenopausal formerly pre‐eclamptic women and premenopausal control women with a history of normotensive pregnancy, which was carried out at the Maastricht University Medical Center in The Netherlands and was approved by the local medical ethics committee. Women who participated in a large ongoing investigation into cardiovascular problems related to pregnancy (the Queen of Hearts study (ClinicalTrials.gov identifier: NCT02347540)) were invited to undergo additional testing. Additionally, women were recruited by their obstetric caregiver and through advertisements. All women provided written informed consent prior to inclusion.
Premenopausal women over 18 years old were considered eligible for the study. Pre‐eclampsia was defined as new‐onset hypertension (systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg) in combination with the development of proteinuria (≥ 300 mg/24 h or a protein‐to‐creatinine ratio of ≥ 0.3) and/or significant end‐organ dysfunction after 20 weeks of pregnancy. The control group had a history of normotensive pregnancy not complicated by pre‐eclampsia, hypertension, intrauterine growth restriction and/or placental abruption. In the control group, the first pregnancy was considered to be the index pregnancy. In the pre‐eclampsia group, the first pre‐eclamptic pregnancy was considered to be the index pregnancy. In both groups, women could participate if they had preterm delivery. Women could not participate if they had delivered within the past 6 months or more than 30 years ago, were diagnosed with chronic hypertension, renal disease or autoimmune disease prior to the index pregnancy, or if they had a contraindication to MRI, such as permanent facial tattoos or an intrauterine device.
Using a standardized format, trained interviewers recorded information about obstetric and medical history, age, educational level and lifestyle habits, such as smoking and alcohol use. Smokers were defined as women who were current smokers or had smoked in the past, while alcohol use was defined as consumption of at least three units of alcohol a week. Height, weight, blood pressure and heart rate were measured on site. Women were diagnosed with hypertension if they had systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg, or used antihypertensive drugs. Body mass index was calculated by dividing body weight (in kg) by the square of height (in m). Body surface area was calculated using the formula of Du Bois and Du Bois 30 . Serum creatinine was determined to assess the glomerular filtration rate, used to define kidney health, to inspect for potential risk related to administration of contrast agent. Blood pressure was measured in a sitting position using a semiautomatic oscillometric device (Dinamap Vital Signs Monitor 1846; Critikon, Tampa, FL, USA) at 3‐min intervals. The median of 11 measurements was used to represent blood pressure. Educational level was recorded using a scale from 1 (primary school) to 8 (academic level). Educational level was categorized further into three subgroups: low (1–3), average (4–6) and high (7–8) following the International Standard Classification of Education.
MRI
Images were obtained using a 7‐Tesla MRI system (Magnetom; Siemens Healthineers, Erlangen, Germany) with a 32‐channel phased‐array head coil (Nova Medical Inc., Wilmington, MA, USA). To improve the image (i.e. transmit field) homogeneity, dielectric pads were placed proximal to the temporal lobe 31 .
To detect BBB leakage, a DCE‐MRI imaging protocol with dual‐time resolution was used, which consisted of two nested T1‐weighted slow and fast sequences 29 , 32 . Before the administration of contrast agent (gadobutrol), precontrast scans of the slow and fast sequences were acquired. Subsequently, contrast agent was injected into the antecubital vein using a power injector (1.0 M gadobutrol, 3 mL; flow rate of 0.3 mL/s), followed by a 20‐mL saline flush.
To visualize the anatomy and to enable conversion of signal intensity changes to gadobutrol concentration, quantitative T1‐mapping was performed prior to the administration of contrast agent. Additionally, whole brain transmit field (B1 +) mapping and a fluid‐attenuated inversion recovery (FLAIR) sequence were performed to improve segmentation. Details of the imaging sequences used are provided in Appendix S1 and Table S1.
An expert radiologist (S.C.G.), blinded to group status, assessed all MRI scans for incidental findings. Additionally, the extent of WM hyperintensities (WMH), defined as WM abnormalities that appear as areas of hyperintense signal on FLAIR images, was investigated using the Fazekas scale 33 . WMH are considered to be of cerebral microvascular origin and are related to hypertension and cerebrovascular risk factors. The Fazekas score provides an overall impression of the presence of WMH in the entire brain.
Image processing
Preprocessing included transmit field correction 34 , bias field correction, skull stripping and gradient distortion correction. The preprocessed images were segmented automatically into subject‐specific gray matter (GM) and WM regions (FreeSurfer software, version 6.0 35 ; Laboratory for Computational Neuroimaging, Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, USA). The segmentation was inspected visually and corrected manually when necessary.
To correct for head displacement, all dynamic images were registered to an averaged precontrast image (FLIRT, FMRIB version 6.0.1; Functional MRI of the Brain Group, Oxford, UK). A participant‐specific signal time course of the blood was obtained in the superior sagittal sinus through manual delineation.
BBB leakage quantification
While the data obtained using the fast DCE‐MRI scan were used only to assess the concentration of gadobutrol in the blood, BBB leakage in the tissue was determined using the slow scan data 29 , 32 . Voxel‐wise pharmacokinetic modeling was applied using the graphical Patlak approach to calculate the BBB leakage rate from the blood plasma to the brain tissue (slope, K i ) and the fractional blood plasma volume in the capillaries (intercept, v p ). Owing to the expected low leakage, the spatial distribution of K i was noisy. Therefore, a histogram approach was used to obtain the corrected K i (magnitude of the K i ) and the fractional leakage volume (v L, the spatial extent of detectable leakage) as the area under the histogram curve after the subtraction of noise 32 . This histogram noise correction approach is explained in Appendix S2 and visualized in Figure S1. The mean BBB measures were calculated for the global GM and WM, deep GM, whole cortical GM and the GM and WM of the frontal, temporal, occipital and parietal lobes. The analysis was performed in MATLAB (R2018b; MathWorks Inc., Natick, MA, USA).
Statistical analysis
Demographic and clinical characteristics were compared between the pre‐eclampsia and control groups using an independent‐sample, two‐sided Student t‐test for continuous variables and the Pearson χ 2 test for categorical variables.
BBB leakage measures of the two groups, including K i , v p and v L , are presented as mean ± SD. Frequency distribution, normal probability plot and the Shapiro–Wilk test were used to check for normality, and Levene's test was used to test for homogeneity of variances. The K i , v L and v p values of the global GM and WM, deep GM, whole cortical GM and the GM and WM of individual brain lobes were assessed between the pre‐eclampsia and control groups using the Mann–Whitney U‐test.
Additionally, odds ratios were calculated to measure the association between higher BBB leakage measures and a history of pre‐eclampsia. BBB leakage was considered high if the values were greater than the mean + 2 SD of those of the control women. Finally, to check whether the observed differences were significant after adjusting for a priori defined potential covariates, including age, hypertension status at MRI and Fazekas score, multivariable linear regression analysis was used to obtain corrected P‐values, and multiple logistic regression was used to estimate adjusted odds ratios; P < 0.05 was considered to indicate statistical significance. Statistical analysis was performed using statistical software SPSS version 23.0 (IBM Corp., Armonk, NY, USA).
RESULTS
Twenty‐two formerly pre‐eclamptic women and 13 parous controls participated in the study (Figure 1). Table 1 presents the demographic and clinical characteristics of the groups. In addition to hypertension, all formerly pre‐eclamptic women developed proteinuria after 20 weeks of the index pregnancy. The average time since the index pregnancy was 6.6 years (range, 1–13 years) in formerly pre‐eclamptic women and 9.0 years (range, 4–15 years) in control women. Fewer women with a history of pre‐eclamptic pregnancy had a subsequent pregnancy. None of the women in the pre‐eclampsia or control group suffered from cardiovascular comorbidities, such as heart attack, stroke, diabetes mellitus or deep vein thrombosis. Formerly pre‐eclamptic women had significantly higher systolic and diastolic blood pressure than did controls. Demographic and clinical characteristics of the formerly pre‐eclamptic women participating in this study were not significantly different from those of all formerly pre‐eclamptic women participating in the larger ongoing Queen of Hearts study (Table S2).
Figure 1.
Flowchart showing inclusion in study population of formerly pre‐eclamptic women and controls with a history of hypertensive pregnancy. COVID‐19, coronavirus disease 2019; MRI, magnetic resonance imaging.
Table 1.
Demographic and clinical characteristics of women who formerly had pre‐eclampsia (PE) and women with a history of normotensive pregnancy (controls)
| Formerly pre‐eclamptic | Controls | ||
|---|---|---|---|
| Characteristic | (n = 22) | (n = 13) | P |
| Demographic | |||
| Age (years) | 37.8 ± 5.4 | 40.8 ± 5.5 | 0.13 |
| Educational level | 0.56 | ||
| Low | 2 (9) | 0 (0) | |
| Average | 8 (36) | 6 (46) | |
| High | 12 (55) | 7 (54) | |
| Index pregnancy | |||
| Early‐onset PE (< 34 weeks) | 13 (59) | NA | |
| HELLP syndrome | 18 (82) | NA | |
| Eclampsia | 2 (9) | NA | |
| Preterm delivery | 17 (77) | 0 (0) | 0.001 |
| Time since index pregnancy (years) | 6.6 ± 3.2 | 9.0 ± 3.7 | 0.05 |
| Subsequent pregnancy | 9 (41) | 10 (77) | 0.02 |
| Early‐onset PE (< 34 weeks) | 0 (0) | NA | |
| HELLP syndrome | 1 (5) | NA | |
| Eclampsia | 0 (0) | NA | |
| Lifestyle | |||
| Smoker | 9 (41) | 5 (38) | 0.75 |
| Alcohol consumer | 14 (64) | 9 (69) | 0.28 |
| Vascular risk profile | |||
| Weight (kg) | 72 ± 11 | 64 ± 8 | 0.03 |
| Height (cm) | 169 ± 5 | 167 ± 6 | 0.24 |
| BMI (kg/m2) | 25.1 ± 4.0 | 22.9 ± 2.8 | 0.08 |
| BSA (m2) | 1.81 ± 0.1 | 1.71 ± 0.1 | 0.03 |
| Hypertension | 6 (27) | 1 (8) | 0.19 |
| SBP (mmHg) | 118 ± 12 | 108 ± 8 | 0.01 |
| DBP (mmHg) | 74 ± 9 | 68 ± 5 | 0.01 |
| MAP (mmHg) | 91 ± 10 | 83 ± 5 | 0.005 |
| Heart rate (bpm) | 66 ± 10 | 67 ± 11 | 0.96 |
| Total cholesterol (mmol/L) | 4.5 ± 1.0 | 4.3 ± 0.7 | 0.42 |
| Glomerular filtration rate (mmol/L/1.73 m2) | 86 ± 12 | 82 ± 11 | 0.33 |
| Fazekas score | 0.60 | ||
| 0 | 14 (64) | 10 (77) | |
| 1 | 8 (36) | 3 (23) | |
| 2 | 0 (0) | 0 (0) | |
| 3 | 0 (0) | 0 (0) | |
Data are presented as mean ± SD or n (%).
Index pregnancy defined as first pregnancy in controls and first PE pregnancy in PE group.
BMI, body mass index; BSA, body surface area; DBP, diastolic blood pressure; MAP, mean arterial pressure; NA, not applicable; SBP, systolic blood pressure.
Examples of leakage maps in a formerly pre‐eclamptic woman and a control woman are presented in Figure 2. The results of the BBB leakage assessment in various brain regions are presented in Table 2. K i in the global WM and GM was significantly higher in formerly pre‐eclamptic women than in controls (P = 0.001 and P = 0.02, respectively). The odds of a high K i were 37 times higher for the global WM and four times higher for the global GM in formerly pre‐eclamptic women compared with controls. While a significantly higher K i was observed in the cortex of formerly pre‐eclamptic women (P = 0.02), K i was not significantly different in the deep GM. Additionally, formerly pre‐eclamptic women had a significantly higher K i in the frontal (P = 0.04) and parietal (P = 0.009) GM and in the frontal (P = 0.001), temporal (P < 0.05) and occipital (P = 0.007) WM (Figure 3). The odds of a high K i in formerly pre‐eclamptic women, when compared with control women, were generally higher in the WM vs GM of different brain lobes. In the same regions, formerly pre‐eclamptic women also had a significantly higher v L than did the control group (Table S3). No significant differences between groups were found in v p (Table S4). A similar trend in BBB leakage differences was found after adjusting for age, hypertension at MRI and Fazekas score (Tables 2, S3 and S4).
Figure 2.
T1‐weighted images (a,c) and leakage rate (K i ) maps (b,d) in a 39‐year‐old formerly pre‐eclamptic woman (a,b) and a 39‐year‐old control woman (c,d). Note the higher K i values (yellow) and higher number of pixels showing any blood–brain barrier leakage, particularly in the frontal lobes, in the formerly pre‐eclamptic woman.
Table 2.
Blood–brain barrier leakage rates in women who formerly had pre‐eclampsia (PE) and in women with a history of normotensive pregnancy (controls)
| Leakage rate ( × 10−7/min) | ||||||
|---|---|---|---|---|---|---|
| Brain region | Women with PE | Controls | Difference | P | OR (95% CI) | aOR (95% CI) |
| Whole cerebrum | ||||||
| White matter | 3.6 ± 3.4 | 0.5 ± 0.9 | 3.1 | 0.001 | 37.4 (3.8–364.6) | 48.3 (3.5–651.2) |
| Gray matter | 13.6 ± 2.7 | 6.0 ± 5.1 | 7.6 | 0.02 | 3.8 (1.1–35.9) | 3.5 (1.1–30.8) |
| Gray matter | ||||||
| Deep | 3.9 ± 3.7 | 2.4 ± 3.6 | 1.5 | 0.26 | 1.1 (0.1–13.5) | 0.7 (0.1–11.0) |
| Cortex | 14.8 ± 14.2 | 7.1 ± 5.2 | 7.7 | 0.02 | 4.2 (1.5–12.4) | 3.9 (1.1–13.2) |
| Lobar gray matter | ||||||
| Frontal | 18.1 ± 16.5 | 7.7 ± 8.2 | 10.4 | 0.04 | 3.4 (1.2–32.6) | 2.8 (1.0–28.7) |
| Temporal | 10.4 ± 1.4 | 5.2 ± 7.0 | 5.2 | 0.06 | 1.6 (0.1–17.2) | 1.0 (0.1–13.1) |
| Parietal | 19.9 ± 17.4 | 6.8 ± 6.5 | 13.1 | 0.009 | 6.9 (1.7–64.0) | 6.6 (1.1–64.4) |
| Occipital | 14.7 ± 13.5 | 7.9 ± 7.8 | 6.8 | 0.12 | 0.8 (0.1–22.1) | 0.9 (0.0–11.6) |
| Lobar white matter | ||||||
| Frontal | 5.6 ± 5.8 | 0.9 ± 1.2 | 4.7 | 0.001 | 10.0 (1.1–92.0) | 10.3 (1.2–102.3) |
| Temporal | 4.3 ± 7.2 | 0.8 ± 1.2 | 3.5 | < 0.05 | 5.1 (1.1–48.0) | 4.1 (1.4–41.0) |
| Parietal | 3.3 ± 4.1 | 1.5 ± 1.9 | 1.8 | 0.09 | 2.4 (0.2–24.8) | 1.8 (0.1–22.1) |
| Occipital | 3.4 ± 4.2 | 0.5 ± 0.6 | 2.9 | 0.007 | 25 (1.2–95.0) | 25.8 (1.2–98.5) |
Data are given as mean ± SD, unless indicated otherwise.
Adjusted odds ratios (aOR) were obtained by correcting for age, hypertension status at magnetic resonance imaging and Fazekas score.
Figure 3.
Average leakage rate in gray matter (a) and white matter (b), according to brain region, in formerly pre‐eclamptic women (
) and in controls with a history of normotensive pregnancy (
). Error bars indicate SD. *Regions with a significantly different leakage rate between groups.
DISCUSSION
BBB leakage was more pronounced throughout the whole cerebrum in formerly pre‐eclamptic women than in parous control women years after the index pregnancy. Higher K i and v L were especially prominent in the frontal and parietal GM, and in the frontal, temporal and occipital WM.
Although in one study BBB permeability was shown to be increased in pregnant and non‐pregnant rats in response to acute hypertension, this difference was not statistically significant 6 . Other studies using pre‐eclampsia models showed increased BBB permeability to small solutes, such as sodium fluorescein (0.47 kDa) and Evans blue (0.96 kDa) 8 , 12 . Increased BBB permeability during gestation was also observed in humans, but the results varied depending on the method of assessment employed 36 , 37 , 38 . BBB leakage of smaller molecules, such as gadobutrol (0.60 kDa), may be more likely when compared with that of larger biomolecules, such as albumin (67 kDa) 39 , 40 . A case report of a 27‐year‐old woman with pre‐eclampsia and posterior reversible encephalopathy syndrome (PRES), described in the case series by Schwartz et al. 41 , 42 , demonstrated BBB breakdown in the occipital lobe and parieto‐occipital junction using a contrast agent. In the present study, we show greater long‐term BBB leakage in women with a history of pre‐eclampsia than that observed in controls with a history of normotensive pregnancy. However, our findings did not match fully the features of PRES, which are often present in the parieto‐occipital region 43 , that were seen in the pregnancy complicated by pre‐eclampsia in studies of Schwartz et al. 41 , 42 . This could be because we investigated the long‐term outcome of pre‐eclampsia instead of the changes in the acute phase. The relationship between the acute and long‐term cerebrovascular outcomes of pre‐eclampsia needs to be investigated in the future.
The increased BBB permeability following pre‐eclampsia could result from temporary elevated blood pressure, as has been suggested by the increase in circulating cells and activation of the endothelium in hypertensive patients 44 . The hypertensive state during a pre‐eclamptic pregnancy could be the cause of BBB leakage. The increased BBB permeability could be related to the breakdown of tight junctions, but circulating factors may also be involved. Although the effect of chronic hypertension on the condition of the BBB is currently unknown, accumulating evidence from animal models suggests that chronic hypertension is accompanied by BBB dysfunction 45 , 46 , 47 . Even though the correction for hypertension status at MRI in our statistical analysis did not significantly alter the results, because of the small number of women with hypertension at the time of MRI (n = 6), we cannot exclude the possibility that long‐term hypertension may contribute to BBB impairment.
WMH have been reported in formerly pre‐eclamptic women up to 5 years after the problematic pregnancy 25 , 26 , 27 . While some studies have suggested a significant difference in terms of WMH between formerly pre‐eclamptic and control women 26 , 27 , other research did not 25 . Wiegman et al. 25 observed WMH in 36.5% of formerly pre‐eclamptic women 5 years after the pregnancy and in 21.3% of control women. Although WMH was more common in pre‐eclamptic women, the difference was not statistically significant. We observed mild WMH (Fazekas score 1) in 36% of cases 6.6 years after the index pregnancy and in 23% of control women 9.0 years after the index pregnancy, similar to the results of Wiegman et al. 25 .
Similarly to our study, Siepmann et al. 48 did not observe a significant difference in the Fazekas score but observed an increase in WMH volume in women with a history of pre‐eclampsia compared with controls. However, the volume of these WMH was very low (∼0.11 cm 3 ) when compared with the hundred times larger volume observed in patients with cSVD 49 . Regardless of pregnancy history, WMH have been linked to BBB disruption and diffuse cerebrovascular endothelial failure 50 . Although we did not observe significantly more WMH in our study, we found an association between greater BBB leakage and pre‐eclampsia, suggesting widespread endothelial dysfunction. The 10 times higher non‐adjusted odds ratio for a high BBB leakage after pre‐eclampsia in the global WM compared with the global GM indicates vulnerability of the generally less vascularized WM 51 , 52 . BBB leakage in WM could be an early marker of WMH and may indicate an increased risk of cerebrovascular disorders, as Hase et al. 53 found evidence of widespread microvascular pathology in neurodegenerative disorders and dementia caused by vascular disease. They observed increased capillary width in the deep WM, suggesting that chronic hypoperfusion induces microvascular modification in the deep WM, which may affect perfusion and result in greater BBB leakage.
A limitation of this study is the relatively small sample size of the formerly pre‐eclamptic group (n = 22) compared with the full study sample (n = 1065). The study characteristics of the formerly pre‐eclamptic women with and those without a MRI scan were not significantly different, which indicates that this sample is representative of the population.
Whether the higher BBB leakage observed is causally related to the impact of pre‐eclampsia on the brain or to the risk profile in which pre‐eclampsia is a vascular stress test cannot be concluded, as other characteristics, such as hypertension and other cardiovascular risk factors, may underlie BBB disruption and pre‐eclampsia. Nevertheless, it is likely that BBB impairment may be involved in (initiating) the pathophysiology and affect brain health years after pre‐eclampsia, as correction for age, hypertension at the time of MRI and Fazekas score did not have a vast impact on the results, indicating that these factors do not underlie the observed differences.
Pre‐eclampsia with severe features, such as eclampsia and HELLP syndrome, has been associated with poorer cognitive performance 54 . These severe features may be associated with greater BBB leakage. However, this should be investigated in the future, as this study aimed to assess the presence and extent of BBB leakage, and its statistical power was too low to assess the impact of pregnancy characteristics.
In conclusion, this study has shown greater BBB leakage in a sample of patients with a history of pre‐eclampsia years after the index pregnancy compared with controls. The increased BBB permeability is subtle but may indicate diffusively spread endothelial dysfunction as a pathophysiological mechanism underlying cerebral injury associated with pre‐eclampsia. Women with a history of pre‐eclampsia may benefit from treatment and prevention strategies aimed at reducing the risk of cerebrovascular disease, such as prolonged neuroprotective treatment, antihypertensive treatment after delivery or lifestyle modifications, such as a healthier diet and physical exercise. Future studies should determine the optimal management strategy for this population.
Supporting information
Appendix S1 Details of magnetic resonance imaging
Appendix S2 Histogram approach for noise correction
Table S1 Pulse sequence parameters
Table S2 Demographic and clinical characteristics of formerly pre‐eclamptic women with and without a cerebral MRI scan participating in the ongoing Queen of Hearts study
Table S3 Fractional leakage volume (v L ) in women with a history of pre‐eclampsia and control women with a history of normal pregnancy
Table S4 Fractional blood plasma volume (v p ) in women with a history of pre‐eclampsia and control women with a history of normal pregnancy
Figure S1 Averaged histograms of the blood–brain barrier leakage rate (K i ) values in the gray and white matter (A1 and B1, respectively) and the histograms corrected for noise by subtraction of the mirrored negative values in the gray and white matter (A2 and B2, respectively). Note the differences in the axis scales in fraction of leaking tissue. The histograms for the formerly pre‐eclamptic women show slightly higher bins for higher K i values (A). After noise correction (B), a larger fraction of voxels with detectable leakage is seen in formerly pre‐eclamptic women.
ACKNOWLEDGMENTS
We thank all other members of the DECONNECT consortium: Eline Kooi and Yentl Brandt from the Maastricht University Medical Center for their help in the study and all members of the Queen of Hearts study group. We also would like to thank Esther Steijvers, Bianca Linssen and Kim Brouwers from Scannexus Ultra‐High Field MRI Center for their assistance. We are also grateful for the aid with source data storage provided by Jos Slenter, Jan Jungerius and Joost de Jong from the Maastricht University Medical Center. Finally, we would like to thank the Brains Unlimited Pioneer Fund (reference S.2016.1.06) and the Stimulation Fund (Stimuleringsfonds FHML) of the Maastricht University for funding this study.
DATA AVAILABILITY STATEMENT
Patient data of this study are available for research purposes in an anonymous format from W.B. and C.G. To access the data a motivated request from a qualified investigator has to be send to the corresponding author.
REFERENCES
- 1. Amaral LM, Cunningham MW Jr, Cornelius DC, LaMarca B. Preeclampsia: long‐term consequences for vascular health. Vasc Health Risk Manag 2015; 11: 403–415. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Brown MC, Best KE, Pearce MS, Waugh J, Robson SC, Bell R. Cardiovascular disease risk in women with pre‐eclampsia: systematic review and meta‐analysis. Eur J Epidemiol 2013; 28: 1–19. [DOI] [PubMed] [Google Scholar]
- 3. Logue OC, George EM, Bidwell GL 3rd. Preeclampsia and the brain: neural control of cardiovascular changes during pregnancy and neurological outcomes of preeclampsia. Clin Sci (Lond) 2016; 130: 1417–1434. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Hammer ES, Cipolla MJ. Cerebrovascular dysfunction in preeclamptic pregnancies. Curr Hypertens Rep 2015; 17: 64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Cipolla MJ. Cerebrovascular function in pregnancy and eclampsia. Hypertension 2007; 50: 14–24. [DOI] [PubMed] [Google Scholar]
- 6. Euser AG, Cipolla MJ. Cerebral blood flow autoregulation and edema formation during pregnancy in anesthetized rats. Hypertension 2007; 49: 334–340. [DOI] [PubMed] [Google Scholar]
- 7. Cipolla MJ, Sweet JG, Chan S. Cerebral vascular adaptation to pregnancy and its role in the neurological complications of eclampsia. J Appl Physiol 2010; 110: 329–339. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Wallace K, Bean C, Bowles T, Spencer S, Randle W, Kyle PB, Shaffery J. Hypertension, anxiety, and blood–brain barrier permeability are increased in postpartum severe preeclampsia/hemolysis, elevated liver enzymes, and low platelet count syndrome rats. Hypertension 2018; 72: 946–954. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Bergman L, Cluver C, Carlberg N, Belfort M, Tolcher MC, Panerai RB, van Veen T. Cerebral perfusion pressure and autoregulation in eclampsia – a case control study. Am J Obstet Gynecol 2021; 225: 185.e1–185.e9. [DOI] [PubMed] [Google Scholar]
- 10. Amburgey OA, Chapman AC, May V, Bernstein IM, Cipolla MJ. Plasma from preeclamptic women increases blood–brain barrier permeability: role of vascular endothelial growth factor signaling. Hypertension 2010; 56: 1003–1008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Bergman L, Acurio J, Leon J, Gatu E, Friis T, Nelander M, Wikström J, Larsson A, Lara E, Aguayo C, Torres‐Vergara P, Wikström A‐K, Escudero C. Preeclampsia and increased permeability over the blood–brain barrier: A role of vascular endothelial growth receptor 2. Am J Hypertens 2021; 34: 73–81. [DOI] [PubMed] [Google Scholar]
- 12. Johnson AC, Tremble SM, Chan S‐L, Moseley J, LaMarca B, Nagle KJ, Cipolla MJ. Magnesium sulfate treatment reverses seizure susceptibility and decreases neuroinflammation in a rat model of severe preeclampsia. PLoS One 2014; 9: e113670. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Cipolla MJ, Kraig RP. Seizures in women with preeclampsia: Mechanisms and management. Fetal Matern Med Rev 2011; 22: 91–108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Warrington JP, Fan F, Murphy SR, Roman RJ, Drummond HA, Granger JP, Ryan MJ. Placental ischemia in pregnant rats impairs cerebral blood flow autoregulation and increases blood–brain barrier permeability. Physiol Rep 2014; 2: e12134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Zlokovic BV. The blood–brain barrier in health and chronic neurodegenerative disorders. Neuron 2008; 57: 178–201. [DOI] [PubMed] [Google Scholar]
- 16. Zhao Z, Nelson AR, Betsholtz C, Zlokovic BV, Angeles L. Establishment and dysfunction of the blood–brain barrier. Cell 2015; 163: 1064–1078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Roes EM, Raijmakers MTM, Schoonenberg M, Wanner N, Peters WHM, Steegers EAP. Physical well‐being in women with a history of severe preeclampsia. J Matern Neonatal Med 2005; 18: 39–45. [DOI] [PubMed] [Google Scholar]
- 18. Elharram M, Dayan N, Kaur A, Landry T, Pilote L. Long‐term cognitive impairment after preeclampsia: A systematic review and meta‐analysis. Obstet Gynecol 2018; 132: 355–365. [DOI] [PubMed] [Google Scholar]
- 19. Heye AK, Culling RD, Valdés Hernández MDC, Thrippleton MJ, Wardlaw JM. Assessment of blood–brain barrier disruption using dynamic contrast‐enhanced MRI. A systematic review. Neuroimage Clin 2014; 6: 262–274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Chassidim Y, Veksler R, Lublinsky S, Pell GS, Friedman A, Shelef I. Quantitative imaging assessment of blood–brain barrier permeability in humans. Fluids Barriers CNS 2013; 10: 9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Sourbron SP, Buckley DL. Classic models for dynamic contrast‐enhanced MRI. NMR Biomed 2013; 26: 1004–1027. [DOI] [PubMed] [Google Scholar]
- 22. Jain R. Measurements of tumor vascular leakiness using DCE in brain tumors: Clinical applications. NMR Biomed 2013; 26: 1042–1049. [DOI] [PubMed] [Google Scholar]
- 23. Jelescu IO, Leppert IR, Narayanan S, Araújo D, Arnold DL, Pike GB. Dual‐temporal resolution dynamic contrast‐enhanced MRI protocol for blood–brain barrier permeability measurement in enhancing multiple sclerosis lesions. J Magn Reson Imaging 2011; 33: 1291–1300. [DOI] [PubMed] [Google Scholar]
- 24. Wong SM, Jansen JFA, Zhang CE, Hoff EI, Staals J, van Oostenbrugge RJ, Backes WH. Blood–brain barrier impairment and hypoperfusion are linked in cerebral small vessel disease. Neurology 2019; 92: e1669–e1677. [DOI] [PubMed] [Google Scholar]
- 25. Wiegman MJ, Zeeman GG, Aukes AM, Bolte AC, Faas MM, Aarnoudse JG, De Groot JC. Regional distribution of cerebral white matter lesions years after preeclampsia and eclampsia. Obstet Gynecol 2014; 123: 790–795. [DOI] [PubMed] [Google Scholar]
- 26. Aukes AM, De Groot JC, Wiegman MJ, Aarnoudse JG, Sanwikarja GS, Zeeman GG. Long‐term cerebral imaging after pre‐eclampsia. BJOG 2012; 119: 1117–1122. [DOI] [PubMed] [Google Scholar]
- 27. Soma‐Pillay P, Suleman FE, Makin JD, Pattinson RC. Cerebral white matter lesions after pre‐eclampsia. Pregnancy Hypertens 2017; 8: 15–20. [DOI] [PubMed] [Google Scholar]
- 28. Schick F. Whole‐body MRI at high field: Technical limits and clinical potential. Eur Radiol 2005; 15: 946–959. [DOI] [PubMed] [Google Scholar]
- 29. Canjels LPW, Jansen JFA, van den Kerkhof M, Alers RJ, Poser BA, Wiggins CJ, Schiffer VMMM, van de ven V, Rouhl RPW, Palm WM, van Oostenbrugge RJ, Aldenkamp AP, Ghossein‐Doha C, Spaanderman MEA, Backes WH. 7T dynamic contrast‐enhanced MRI for the detection of subtle blood–brain barrier leakage. J Neuroimaging 2021; 31: 902–911. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Du Bois D, Du Bois EF. A formula to estimate the approximate surface area if height and weight be known. 1916. Nutrition 1989; 5: 303–311. [PubMed] [Google Scholar]
- 31. Teeuwisse WM, Brink WM, Webb AG. Quantitative assessment of the effects of high‐permittivity pads in 7 Tesla MRI of the brain. Magn Reson Med 2012; 67: 1285–1293. [DOI] [PubMed] [Google Scholar]
- 32. van de Haar HJ, Burgmans S, Jansen JFA, van Osch MJP, van Buchem MA, Muller M, Hofman PAM, Verhey FRJ, Backes WH. Blood–Brain Barrier leakage in patients with early Alzheimer disease. Radiology 2016; 281: 527–535. [DOI] [PubMed] [Google Scholar]
- 33. Fazekas F, Kleinert R, Offenbacher H, Schmidt R, Kleinert G, Payer F, Radner H, Lechner H. Pathologic correlates of incidental MRI white matter signal hyperintensities. Neurology 1993; 43: 1683–1689. [DOI] [PubMed] [Google Scholar]
- 34. Haast RAM, Ivanov D, Uludağ K. The impact of B1 + correction on MP2RAGE cortical T1 and apparent cortical thickness at 7T. Hum Brain Mapp 2018; 39: 2412–2425. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Fischl B, Salat DH, Busa E, Albert M, Dieterich M, Haselgrove C, Van Der Kouwe A, Killiany R, Kennedy D, Klaveness S, Montillo A, Makris N, Rosen B, Dale AM. Whole brain segmentation: Automated labeling of neuroanatomical structures in the human brain. Neuron 2002; 33: 341–355. [DOI] [PubMed] [Google Scholar]
- 36. Burwick RM, Togioka BM, Speranza RJ, Gaffney JE, Roberts VHJ, Frias AE, Rincón M. Assessment of blood–brain barrier integrity and neuroinflammation in preeclampsia. Am J Obstet Gynecol 2019; 221: 269.e1–269.e8. [DOI] [PubMed] [Google Scholar]
- 37. Andersson M, Oras J, Thörn SE, Karlsson O, Kälebo P, Zetterberg H, Blennow K, Bergman L. Signs of neuroaxonal injury in preeclampsia – A case control study. PLoS One 2021; 16: e0246786. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Bergman L, Hastie R, Zetterberg H, Blennow K, Schell S, Langenegger E, Moodley A, Walker S, Tong S, Cluver C. Evidence of neuroinflammation and blood–brain barrier disruption in women with preeclampsia and eclampsia. Cells 2021; 10: 3045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Ciampa E, Li Y, Dillon S, Lecarpentier E, Sorabella L, Libermann TA, Karumanchi SA, Hess PE. Cerebrospinal fluid protein changes in preeclampsia. Hypertension 2018; 72: 219–226. [DOI] [PubMed] [Google Scholar]
- 40. van den Berg CB, Duvekot JJ, Güzel C, Hansonn SR, de Leeuw TG, Steegers EAP, Versendaal J, Luider TM, Stoop MP. Elevated levels of protein AMBP in cerebrospinal fluid of women with preeclampsia compared to normotensive pregnant women. Proteomics Clin Appl 2017; 11: 1–2. [DOI] [PubMed] [Google Scholar]
- 41. Schwartz RB, Jones KM, Kalina P, Bajakian RL, Mantello MT, Garada B, Holman BL. Hypertensive encephalopathy: findings on CT, MR imaging, and SPECT imaging in 14 Cases. AJR Am J Roentgenol 1992; 159: 379–383. [DOI] [PubMed] [Google Scholar]
- 42. Schwartz RB, Feske SK, Polak JF, DeGirolami U, Iaia A, Beckner KM, Bravo SM, Klufas RA, Chai RYC, Repke JT. Preeclampsia–eclampsia: Clinical and neuroradiographic correlates and insights into the pathogenesis of hypertensive encephalopathy. Radiology 2000; 217: 371–376. [DOI] [PubMed] [Google Scholar]
- 43. Parasher A, Jhamb R. Posterior reversible encephalopathy syndrome (PRES): presentation, diagnosis and treatment. Postgr Med J 2020; 96: 623–628. [DOI] [PubMed] [Google Scholar]
- 44. Meissner A. Hypertension and the brain : A risk factor for more than heart disease. Cerebrovasc Dis 2016; 42: 255–262. [DOI] [PubMed] [Google Scholar]
- 45. Kruyer A, Soplop N, Strickland S, Norris EH. Chronic hypertension leads to neurodegeneration in the TgSwDI mouse model of Alzheimer's disease. Hypertension 2015; 66: 175–182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Özkan E, Çetin‐Taş Y, Şekerdağ E, Kızılırmak AB, Taş A, Yıldız E, Yapıcı‐Eser H, Karahüseyinoğlu S, Zeybel M, Gürsoy‐Özdemir Y. Blood–brain barrier leakage and perivascular collagen accumulation precede microvessel rarefaction and memory impairment in a chronic hypertension animal model. Metab Brain Dis 2021; 36: 2553–2566. [DOI] [PubMed] [Google Scholar]
- 47. Buttler L, Jordão MT, Fragas MG, Ruggeri A, Ceroni A, Michelini LC. Maintenance of blood–brain barrier integrity in hypertension: A novel benefit of exercise training for autonomic control. Front Physiol 2017; 8: 1048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Siepmann T, Boardman H, Bilderbeck A, Griffanti L, Kenworthy Y, Zwager C, McKean D, Francis J, Neubauer S, Yu GZ, Lewandowski AJ, Sverrisdottir YB, Leeson P. Long‐term cerebral white and gray matter changes after preeclampsia. Neurology 2017; 88: 1256–1264. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Papma JM, de Groot M, de Koning I, Mattace‐Raso FU, van der Lugt A, Vernooij MW, Niessen WJ, van Swieten JC, Koudstaal PJ, Prins ND, Smits M. Cerebral small vessel disease affects white matter microstructure in mild cognitive impairment. Hum Brain Mapp 2014; 35: 2836–2851. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Wardlaw JM, Smith C. Mechanisms underlying sporadic cerebral small vessel disease: insights from neuroimaging. Lancet Neurol 2013; 12: 483–497. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. Leenders KL, Perani D, Lammertsma AA, Heather JD, Buckingham P, Healy MJ, Gibbs JM, Wise RJ, Hatazawa J, Herold S. Cerebral blood flow, blood volume and oxygen utilization. Normal values and effect of age. Brain 1990; 113: 27–47. [DOI] [PubMed] [Google Scholar]
- 52. Lierse W, Horstmann E. Quantitative anatomy of the cerebral vascular bed with especial emphasis on homogeneity and inhomogeneity in small ports of the gray and white matter. Acta Neurol Scand Suppl 1965; 14: 15–19. [DOI] [PubMed] [Google Scholar]
- 53. Hase Y, Ding R, Harrison G, Hawthorne E, King A, Gettings S, Platten C, Stevenson W, Craggs LJL, Kalaria RN. White matter capillaries in vascular and neurodegenerative dementias. Acta Neuropathol Commun 2019; 7: 16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54. Bergman L, Thorgeirsdottir L, Elden H, Hesselman S, Schell S, Ahlm E, Aukes A, Cluver C. Cognitive impairment in preeclampsia complicated by eclampsia and pulmonary edema after delivery. Acta Obstet Gynecol Scand 2021; 100: 1280–1287. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Appendix S1 Details of magnetic resonance imaging
Appendix S2 Histogram approach for noise correction
Table S1 Pulse sequence parameters
Table S2 Demographic and clinical characteristics of formerly pre‐eclamptic women with and without a cerebral MRI scan participating in the ongoing Queen of Hearts study
Table S3 Fractional leakage volume (v L ) in women with a history of pre‐eclampsia and control women with a history of normal pregnancy
Table S4 Fractional blood plasma volume (v p ) in women with a history of pre‐eclampsia and control women with a history of normal pregnancy
Figure S1 Averaged histograms of the blood–brain barrier leakage rate (K i ) values in the gray and white matter (A1 and B1, respectively) and the histograms corrected for noise by subtraction of the mirrored negative values in the gray and white matter (A2 and B2, respectively). Note the differences in the axis scales in fraction of leaking tissue. The histograms for the formerly pre‐eclamptic women show slightly higher bins for higher K i values (A). After noise correction (B), a larger fraction of voxels with detectable leakage is seen in formerly pre‐eclamptic women.
Data Availability Statement
Patient data of this study are available for research purposes in an anonymous format from W.B. and C.G. To access the data a motivated request from a qualified investigator has to be send to the corresponding author.