Background
Preeclampsia complicates approximately 5% of all pregnancies. When pulmonary edema occurs, it accounts for 50% of preeclampsia-related mortality. Currently, there is no consensus on the degree to which left ventricular systolic dysfunction contributes to the development of pulmonary edema.
Objective
This study aimed to use cardiac magnetic resonance imaging to detect subtle changes in left ventricular systolic function and evidence of acute left ventricular dysfunction (through tissue characterization) in women with preeclampsia complicated by pulmonary edema compared with both preeclamptic and normotensive controls.
Study Design
Cases were postpartum women aged ≥18 years presenting with preeclampsia complicated by pulmonary edema. Of note, 2 control groups were recruited: women with preeclampsia without pulmonary edema and women with normotensive pregnancies. All women underwent echocardiography and 1.5T cardiac magnetic resonance imaging with native T1 and T2 mapping. Gadolinium contrast was administered to cases only. Because of small sample sizes, a nonparametric test (Kruskal-Wallis) with pairwise posthoc analysis using Bonferroni correction was used to compare the differences between the groups. Cardiac magnetic resonance images were interpreted by 2 independent reporters. The intraclass correlation coefficient was calculated to assess interobserver reliability.
Results
Here, 20 women with preeclampsia complicated by pulmonary edema, 13 women with preeclampsia (5 with severe features and 8 without severe features), and 6 normotensive controls were recruited. There was no difference in the baseline characteristics between groups apart from the expected differences in blood pressure. Left atrial sizes were similar across all groups. Women with preeclampsia complicated by pulmonary edema had increased left ventricular mass ( P =.01) but had normal systolic function compared with the normotensive controls. Furthermore, they had elevated native T1 values ( P =.025) and a trend toward elevated T2 values ( P =.07) in the absence of late gadolinium enhancement consistent with myocardial edema. Moreover, myocardial edema was present in all women with eclampsia or hemolysis, elevated liver enzymes, and low platelet count. Women with preeclampsia without severe features had similar findings to the normotensive controls. All cardiac magnetic resonance imaging measurements showed a very high level of interobserver correlation.
Conclusion
This study focused on cardiac magnetic resonance imaging in women with preeclampsia complicated by pulmonary edema, eclampsia, and hemolysis, elevated liver enzymes, and low platelet count. We have demonstrated normal systolic function with myocardial edema in women with preeclampsia with these severe features. These findings implicate an acute myocardial process as part of this clinical syndrome. The pathogenesis of myocardial edema and its relationship to pulmonary edema require further elucidation. With normal left atrial sizes, any hemodynamic component must be acute.
Why was this study conducted?
The mechanisms underlying the development of pulmonary edema in preeclampsia are not well understood, and the literature is conflicting. We evaluated left ventricular (LV) function and tissue characterization in women with preeclampsia with severe features and controls using cardiac magnetic resonance imaging (MRI).
Key findings
LV systolic dysfunction did not play a role in the development of pulmonary edema. Evidence of myocardial edema was found in women with preeclampsia complicated by pulmonary edema; hemolysis, elevated liver enzymes, and low platelet count; and eclampsia.
What does this add to what is known?
Systolic dysfunction was not the mechanism underlying pulmonary edema in preeclampsia. The presence of myocardial edema in preeclampsia with severe features without pulmonary edema suggested more widespread subclinical myocardial involvement than previously suspected. The absence of left atrial enlargement suggested that any hemodynamic cause for pulmonary edema is acute.
Introduction
Preeclampsia complicates 3% to 5% of all pregnancies. Pulmonary edema is a defining feature of preeclampsia with severe features and accounts for up to 50% of preeclampsia-related deaths. , In nonpregnant patients with hypertension, diastolic dysfunction has been identified as the usual cause of acute pulmonary edema. Similarly, in women with preeclampsia, the prevailing opinion seems to favor a composite of diastolic dysfunction, increased afterload (because of endothelial dysfunction and vasoconstriction), and capillary leak with interstitial edema. ,
In studies of women with preeclampsia who have undergone echocardiography, some have reported normal systolic function accompanied by diastolic dysfunction, whereas others have shown systolic dysfunction to a greater or lesser degree. Most studies did not assess women with pulmonary edema as a distinct group. There are limited data on the echocardiographic findings in women with preeclampsia complicated by pulmonary edema.
Cardiac magnetic resonance imaging (MRI) has become the reference standard for quantifying chamber size and ejection fraction. Despite the favorable risk-benefit ratio of cardiac MRI in pregnancy, it remains an underutilized tool. Of note, 2 small studies have reported cardiac MRI findings in women with preeclampsia, and no woman with preeclampsia complicated by pulmonary edema was included. , Both studies primarily reported left ventricular (LV) volumes and masses, and 1 study reported on short-tau inversion recovery (STIR) imaging findings. With its ability to characterize myocardial pathology, for instance, through edema imaging (STIR, native T1 and T2 mapping), cardiac MRI offers the opportunity to determine the pathophysiological basis for the anatomic abnormalities detected. There is no report detailing native (pregadolinium contrast) tissue mapping.
Considering the paucity of data and the overall poor understanding of the mechanisms and pathophysiology underlying pulmonary edema in women with preeclampsia, we set out to use cardiac MRI to assess for subtle changes in LV systolic function in women with preeclampsia complicated by pulmonary edema. Furthermore, we utilized tissue characterization to assess for evidence of acute myocardial edema and altered gadolinium kinetics as markers of acute ventricular pathology leading to dysfunction. By comparing these findings in women with preeclampsia complicated by pulmonary edema with those in women with preeclampsia without pulmonary edema and normotensive controls, we aimed to advance the understanding of the pathologic mechanisms underlying pulmonary edema in preeclampsia.
Materials and Methods
This prospective observational study (S17/10/260) was approved by the human research ethics committee of the University of Stellenbosch, Cape Town, South Africa. Furthermore, these women were included in the Preeclampsia Obstetric Adverse Event biobank.
Study population
We included postpartum women who delivered at Tygerberg Hospital in Cape Town, South Africa. Cases were women with preeclampsia complicated by pulmonary edema. These cases were compared with cases of women with preeclampsia without pulmonary edema and normotensive controls. Women with other target organ dysfunction, such as eclampsia and hemolysis, elevated liver enzymes, and low platelet count (HELLP syndrome), were eligible for inclusion in the preeclampsia control group.
We included women aged ≥18 years without secondary hypertension. The exclusion criteria included non–preeclamptic-related pulmonary edema, an alternative cause for respiratory failure, and claustrophobia.
Pulmonary edema was diagnosed in the presence of respiratory failure necessitating respiratory support not because of infective or embolic causes, the presence of bibasal fine inspiratory crackles on auscultation, which responded to diuresis or positive pressure ventilation and/or compatible radiologic features on chest x-ray or computed tomography scanning.
Women were classified using the International Society for the Study of Hypertension in Pregnancy definition of preeclampsia. Moreover, significant proteinuria was defined as a urine protein-to-creatinine ratio (PCR) of ≥30 mg/mmol (0.3 mg/mg). A 24-hour urine PCR of ≥0.3 g or urine dipstick >1+ (if a PCR was not available) was required to make the diagnosis. Women with only hypertension and proteinuria were classified as having preeclampsia without severe features. We did not include severe hypertension as a feature of severe disease. Women with acute renal injury, liver dysfunction, neurologic features (including eclampsia), hemolysis, or thrombocytopenia were classified as having preeclampsia with severe features. A woman was considered normotensive if she had no documented systolic blood pressure (BP) of >139 mm Hg or diastolic BP of >89 mm Hg during her pregnancy until discharge after birth. Body surface area (BSA) and body mass index (BMI) were calculated using postpartum weight and height. BSA was calculated using the DuBois formula (0.0247 × height [m] 0.725 × weight [kg] 0.425 ). Medical therapy was at the discretion of the treating physician.
Women were identified in the high care unit, labor ward, and postpartum ward by the investigators and research nurses. Folders were reviewed to find the most appropriate candidates. Cases and controls were concurrently recruited, but women with preeclampsia with pulmonary edema were prioritized for inclusion. After signing the informed consent, all women underwent cardiac MRI. Baseline data were obtained by interview and extraction from medical records and entered and stored using Research Electronic Data Capture tools (Siemens AG Munich, Germany) hosted at the University of Stellenbosch. , Electronic data were double-checked for accuracy and cross-referenced with original data collection forms collected by research midwives.
Magnetic resonance imaging assessment
Cardiac MRI was performed following current guideline recommendations. A 1.5T cardiac MRI (Siemens Aera, REDcap Nashville,TN) was performed as soon as cardiac MRI could be tolerated. All cases with pulmonary edema had gadolinium contrast (Gadavist Bayer Healthcare Pharmaceuticals Inc. Leverkeusen, Germany) administered at the standard recommended dosage (0.2 mL/kg). Controls had a cardiac MRI without gadolinium. The analysis of the cardiac MRI data was performed using CVI 42 (Circle Cardiovascular Imaging Inc, Calgary, Canada). We determined LV volumes, mass, and functional parameters. We assessed edema using STIR sequences. Furthermore, native T1 and T2 tissue characterizations were performed. Further detail on the methodology can be found in the Supplemental Material section. All cardiac MRIs were read by a single reporter (L.H.J.) who was not blinded to the participant group. To reduce observer bias, the first 27 of 39 MRIs were independently reported by a second reporter (A.S.H.), who was blinded to the participant group.
Statistical methods
Data were analyzed using the SPSS Statistics software (version 27; IBM SPSS Statistics, IBM Corporation 2020, Chicago, IL). Participant characteristics and cardiac MRI variables were expressed as means with standard deviations, medians with interquartile ranges, and numbers with percentages, according to order of the variables and distribution of data. Categorical variables were compared using cross-tabulation and the chi-square test. Because of small sample sizes, a nonparametric test (independent-samples Kruskal-Wallis test) with pairwise post hoc analysis using Bonferroni correction was used to compare the differences between the groups with normal distribution. The Mann-Whitney U test was used to compare the means of groups without normal distribution. Correlation between continuous variables was evaluated using the Pearson correlation coefficient. A P value of <.05 was used to ascribe statistical significance. The intraclass correlation coefficient was calculated to assess interobserver reliability for the cardiac MRI reports for each of the measured variables.
Results
A total of 42 postpartum women underwent cardiac MRI. Of note, 3 women were excluded from the final analysis. Moreover, 1 woman with pulmonary edema was excluded because of mapping-related artifact, which precluded accurate assessment of the T1 and T2 sequences. Furthermore, 2 normotensive postpartum controls were excluded. Of the 2 women, one had absent native T2 mapping data, and the other had non–preeclampsia-related myocardial abnormalities on MRI. Overall, 39 women were included in the final analysis. Of those women, 20 had preeclampsia complicated by pulmonary edema, and 13 women had preeclampsia without pulmonary edema, 5 of whom had preeclampsia with severe features (4 with eclampsia and 1 with HELLP syndrome). Moreover, 6 normotensive postpartum women were included. Of the 4 women with chronic hypertension were included in the study. Of the 4 women, 3 were in the pulmonary edema group, and 1 was included in the preeclampsia without severe features group.
There was no significant difference between women with preeclampsia complicated by pulmonary edema and controls in terms of mean age, gestation, gravidity and parity, BSA, BMI, and heart rate at the time of inclusion ( Table 1 ). As expected, there was a statistically significant difference in systolic and diastolic BPs between the groups. Patients with preeclampsia complicated by pulmonary edema had significantly higher BPs than both preeclamptic and nonpreeclamptic controls. When the control cases with preeclampsia were divided into those with and without severe features, the difference in BP persisted compared with normotensive controls, but there was no other significant difference in the baseline characteristics ( Table 2 ). The use of antihypertensive medication, namely, nifedipine, methyldopa, and intravenous labetalol, was similar for all women with preeclampsia. Women with pulmonary edema were more likely to receive intravenous furosemide. Women with any form of preeclampsia were more likely to deliver via cesarean delivery, whereas normotensive controls were more likely to deliver vaginally (X 2 [9, n=39] = 23.18; P =.006)
| Characteristic | Preeclampsia with pulmonary edema (n=20) | SD or % | Preeclampsia without pulmonary edema(n=13) | SD or % | Normotensive controls (n=6) | SD or % | P value |
|---|---|---|---|---|---|---|---|
| Demographics | |||||||
| Age | 28.3 | 5.9 | 27.6 | 6.0 | 28.0 | 5.8 | .96 |
| Gravidity | 2.5 | 1.3 | 2.4 | 1.5 | 2.2 | 1.2 | .87 |
| Parity | 1.2 | 1.1 | 1.2 | 1.5 | 1.5 | 0.8 | .83 |
| Gestation (wk) | 32.2 | 5.5 | 34.8 | 5.0 | 35.2 | 6.1 | .243 |
| CD | 16 | 80% | 10 | 77% | 2 | 33% | <.001 |
| BMI (kg/m 2 ) | 34.1 | 12.7 | 35.9 | 7.9 | 34.4 | 9.6 | .58 |
| BSA (m 2 ) | 1.9 | 0.4 | 2.1 | 0.2 | 1.9 | 0.2 | .269 |
| Heart rate (bpm) | 95.0 | 11.0 | 95.5 | 16.0 | 86.8 | 18.5 | .475 |
| Systolic BP (mm Hg) | 189.4 | 24.9 | 163.9 | 18.1 | 116.0 | 8.7 | <.001 |
| Diastolic BP (mm Hg) | 110.6 | 15.5 | 103.6 | 20.6 | 62.3 | 5.2 | <.001 |
| Imaging findings | |||||||
| LVEF (%) | 62.8 | 10.9 | 68.5 | 4.4 | 58.2 | 7.7 | .027 |
| LVEDV (mL) | 149.6 | 34.7 | 145.1 | 29.3 | 141.3 | 18.3 | .942 |
| LVESV (mL) | 55.3 | 22.2 | 45.3 | 9.1 | 58.9 | 13.6 | .194 |
| LVMI (g/m 2 ) | 77.4 | 19.9 | 67.3 | 15.0 | 56.0 | 7.6 | .01 |
| LA area | 22.8 | 4.3 | 21.4 | 3.2 | 22.3 | 4.1 | .625 |
| LAVI (mL/m 2 ) | 38.9 | 9.2 | 32.7 | 6.9 | 34.8 | 7.3 | .207 |
| Basal STIR | 1.1 | 0.6 | 0.9 | 0.4 | 1.2 | 0.4 | .377 |
| Middle STIR | 1.2 | 0.5 | 1.1 | 0.2 | 1.4 | 0.2 | .02 |
| Apical STIR | 1.1 | 0.3 | 1.1 | 0.3 | 1.4 | 0.2 | .05 |
| Global T1 (ms) | 1082.4 | 52.7 | 1032.3 | 52.5 | 1037.3 | 19.8 | .025 |
| Global T2 (ms) | 55.5 | 7.0 | 53.5 | 5.1 | 49.2 | 1.7 | .07 |
| Time to MRI (d) | 5.4 | 3.0 | 1.5 | 1.1 | 1.7 | 0.8 | <.001 |
| Characteristic | Preeclampsia with pulmonary edema (n=20) | SD or % | Eclampsia or HELLP syndrome (n=5) | SD or % | Preeclampsia without severe features (n=8) | SD or% | Normotensive controls (n=6) | SD | P value |
|---|---|---|---|---|---|---|---|---|---|
| Demographics | |||||||||
| Age (y) | 28.3 | 5.9 | 26.0 | 7.0 | 28.6 | 5.5 | 28 | 5.8 | .93 |
| Gravidity | 2.5 | 1.3 | 3.0 | 1.9 | 2.0 | 1.2 | 2.2 | 1.2 | .583 |
| Parity | 1.2 | 1.1 | 1.8 | 1.8 | 0.8 | 1.2 | 1.5 | 0.8 | .435 |
| Gestation (wk) | 32.2 | 5.5 | 34.3 | 5.7 | 35.2 | 4.9 | 35.2 | 6.1 | .401 |
| CD | 16 | 80% | 4 | 80% | 6 | 75% | 2 | 33% | .006 |
| BMI (kg/m 2 ) | 34.1 | 12.7 | 38.1 | 11.1 | 34.5 | 5.6 | 34.4 | 9.6 | .747 |
| BSA (m 2 ) | 1.9 | 0.4 | 2.1 | 0.3 | 2.0 | 0.1 | 1.9 | 0.2 | .453 |
| Heart rate (bpm) | 94.6 | 10.9 | 93.0 | 16.5 | 97.0 | 16.6 | 86.8 | 18.5 | .545 |
| Systolic BP (mm Hg) | 189.4 | 24.9 | 172.4 | 16.2 | 158.6 | 18.1 | 116.0 | 8.7 | <.001 |
| Diastolic BP (mm Hg) | 110.6 | 15.5 | 122.0 | 14.8 | 92.1 | 14.5 | 62.3 | 5.2 | <.001 |
| Imaging findings | |||||||||
| LVEDV (mL) | 149.6 | 34.7 | 170.1 | 26.1 | 129.5 | 18.7 | 141.3 | 18.3 | .123 |
| LVESV (mL) | 55.3 | 22.2 | 47.9 | 9.7 | 43.7 | 8.9 | 58.9 | 13.6 | .303 |
| LVEF (%) | 62.8 | 10.9 | 71.8 | 4.1 | 66.4 | 3.4 | 58.2 | 7.7 | .02 |
| LVMI (g/m 2 ) | 77.4 | 19.9 | 79.4 | 12.0 | 59.7 | 11.4 | 56.0 | 7.6 | .002 |
| LA area | 22.8 | 4.3 | 23.5 | 3.1 | 20.0 | 2.5 | 22.3 | 4.1 | .274 |
| LAVI (mL/m 2 ) | 38.9 | 9.2 | 34.7 | 3.7 | 31.5 | 8.3 | 34.8 | 7.3 | .322 |
| Basal STIR | 1.1 | 0.6 | 0.8 | 0.5 | 1.0 | 0.3 | 1.2 | 0.4 | .372 |
| Middle STIR | 1.2 | 0.5 | 1.1 | 0.3 | 1.1 | 0.2 | 1.4 | 0.2 | .049 |
| Apical STIR | 1.1 | 0.3 | 1.0 | 0.3 | 1.2 | 0.3 | 1.4 | 0.2 | .094 |
| Global T1 (ms) | 1082.4 | 52.7 | 1081.8 | 43.6 | 1001.4 | 28.3 | 1037.3 | 19.8 | .002 |
| Global T2 (ms) | 55.5 | 7.0 | 55.4 | 6.8 | 52.4 | 3.9 | 49.2 | 1.7 | .114 |
| Time to MRI (d) | 5.4 | 3.0 | 2.2 | 1.6 | 1.1 | 0.4 | 1.7 | 0.8 | <.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
