Circulating vascular cell adhesion molecule–1 in pre-eclampsia, gestational hypertension, and normal pregnancy: Evidence of selective dysregulation of vascular cell adhesion molecule–1 homeostasis in pre-eclampsia

Circulating vascular cell adhesion molecule–1 in pre-eclampsia, gestational hypertension, and normal pregnancy: Evidence of selective dysregulation of vascular cell adhesion molecule–1 homeostasis in pre-eclampsia John R. Higgins, MD,a Aikaterina Papayianni, MD,b Hugh R. Brady, MD,c Michael R.N. Darling, MD,a and Joseph J. Walshe, MDa Dublin, Ireland, and Boston, Massachusetts OBJECTIVE: Our purpose was to investigate circulating levels of vascular cell adhesion molecule–1 in the peripheral and uteroplacental circulations during normotensive and hypertensive pregnancies. STUDY DESIGN: This prospective observational study involved 2 patient groups. Group 1 consisted of 22 women with pre-eclampsia and 30 normotensive women followed up longitudinally through pregnancy and post partum. There were an additional 13 women with established gestational hypertension. Group 2 consisted of 20 women with established pre-eclampsia and 19 normotensive control subjects undergoing cesarean delivery. Plasma levels of vascular cell adhesion molecule–1 were measured in blood drawn from the antecubital vein (group 1) and from both the antecubital and uterine veins (group 2). Data were analyzed by analysis of variance. RESULTS: In group 1 vascular cell adhesion molecule–1 levels did not change significantly throughout normal pregnancy and post partum. Women with established pre-eclampsia had increased vascular cell adhesion molecule–1 levels compared with the normotensive pregnancy group (P = .01). Vascular cell adhesion molecule–1 levels were not elevated in women with established gestational hypertension. In group 2 significantly higher levels of vascular cell adhesion molecule–1 were detected in the uteroplacental (P < .0001) and peripheral (P < .0001) circulations of pre-eclamptic women by comparison with normotensive women. In the pre-eclamptic group there was a tendency toward higher vascular cell adhesion molecule–1 levels in the peripheral circulation than in the uteroplacental circulation (P = .06). In contrast to vascular cell adhesion molecule–1, circulating levels of E-selectin and intercellular adhesion molecule–1, other major leukocyte adhesion molecules expressed by the endothelium, were not different in pre-eclamptic and normotensive pregnancies. CONCLUSION: Established pre-eclampsia is characterized by selective dysregulation of vascular cell adhesion molecule–1 homeostasis. This event is not an early preclinical feature of pre-eclampsia, does not persist post partum, is not a feature of nonproteinuric gestational hypertension, and is not observed with other major leukocyte adhesion molecules. Induction of vascular cell adhesion molecule–1 expression in pre-eclampsia may contribute to leukocyte-mediated tissue injury in this condition or may reflect perturbation of other, previously unrecognized, functions of this molecule in pregnancy. (Am J Obstet Gynecol 1998;179:464-9.)

Key words: Pre-eclampsia, leukocyte activation, cell adhesion molecules

Endothelial activation is a key process in the pathogenesis of pre-eclampsia.1 The primary pathologic event appears to be failure of the normal trophoblastic invasion From the Rotunda Hospital,a the Brockton-West Roxbury Veterans Affairs Medical Centre and the Brigham and Women’s Hospital, Harvard Medical School,b and the Department of Medicine and Therapeutics, Mater Hospital, University College.c Funded by research grants from the Health Research Board, the Jervis Street Charitable Infirmary Charitable Trust, the Friends of the Rotunda, and the Irish Perinatal Society. See also article by Higgins et al on pages 520-6. Received for publication August 11, 1997; revised January 12, 1998; accepted January 28, 1998. Reprint requests: J.R. Higgins, MD, Department of Perinatal Medicine, Royal Women’s Hospital, Carlton, Victoria 3053, Australia. 6/1/89128

464

of the spiral arteries. It has further been proposed that perturbed trophoblastic invasion leads to placental ischemia and the subsequent release of an unknown factor or factors that cause widespread endothelial damage.1 Well-recognized consequences of endothelial activation in human disease include abnormal vascular tone, platelet activation, and leukocyte activation. The first 2 of these abnormalities have been well defined in preeclampsia, but the role of leukocyte activation is poorly understood. Recent evidence suggests that leukocytes are potential mediators of tissue injury in this syndrome.2 Leukocyte– endothelial cell adhesion plays a central role in leukocyte homing to sites of extravascular inflammation and is sup-

Volume 179, Number 2 Am J Obstet Gynecol

ported by binding of specific adhesion molecules on leukocytes with endothelial cell ligands.3 Leukocyte adhesion molecules can be cleaved or shed from the cell surface and can be detected circulating in blood, where they are also potential modulators of leukocyte–endothelial cell interactions.4 Adhesion molecules involved in leukocyte adhesion include immunoglobulin-like molecules such as vascular cell adhesion molecule–1, intercellular adhesion molecule–1, and E-selectin. Patterns of expression of leukocyte adhesion molecules in pre-eclampsia are still being defined. Results for intercellular adhesion molecule–1 and E-selectin have been conflicting, with some studies reporting increased circulating levels in pre-eclampsia5, 6 and others reporting no change.7, 8 In contrast, all studies to date have reported an increase in circulating vascular cell adhesion molecule–1 levels.5-8 The aims of our study were to explore the latter observation further by (1) longitudinally studying vascular cell adhesion molecule–1 expression in normal pregnancy and pre-eclampsia to define the temporal relationship of vascular cell adhesion molecule–1 expression with the development of pre-eclampsia, (2) determining whether this abnormality is specific for pre-eclampsia or a general consequence of the hemodynamic stress of increased blood pressure on the vascular endothelium, and (3) probing the source of vascular cell adhesion molecule–1 and the possible contribution of renal failure to its accumulation and recirculation. Methods Group 1 patients. A total of 1001 healthy primigravid women who had been recruited for a study on the predictive value of ambulatory monitoring in pregnancy agreed to have blood samples taken between 18 and 24 weeks’ gestation.9 All women were normotensive at this sampling. Thirty women were recruited as control subjects for the longitudinal study and had additional blood samples drawn at 28 to 32 weeks’ gestation, ≥33 weeks’ gestation, and 6 weeks post partum. When possible any woman within the original cohort of 1001 women in whom pre-eclampsia subsequently developed had blood sampling repeated when the condition developed and again at 6 weeks post partum. A separate group of 13 women with established gestational hypertension was also recruited. Group 2 patients. Twenty women with confirmed preeclampsia were recruited at the time of cesarean delivery. Nineteen normotensive women undergoing elective cesarean delivery were recruited as control subjects. After delivery of the infant but before delivery of the placenta a blood sample was taken from a uterine vein draining the placental site. A blood sample was simultaneously taken from the antecubital vein. None of these women were in established labor. Specific exclusion criteria for

Higgins et al 465

all women involved in the study included a history of hypertension, renal disease, cardiac disease, or diabetes mellitus. Interim analysis was performed on the first 12 women with pre-eclampsia and 8 normotensive control subjects; plasma levels of vascular cell adhesion molecule–1, intercellular adhesion molecule–1, and E-selectin were all measured. On the basis of these results only vascular cell adhesion molecule–1 levels were measured in the full cohort. Blood sampling. Patients in group 1 had blood samples taken from the antecubital vein with a Monovette system of blood collection (Sarstedt Ireland Ltd, Numbrecht, Germany), with minimum venous stasis. Four milliliters of blood was taken into a lithium heparin container. Blood samples in group 2 were taken into dry plastic syringes and immediately transferred into the heparinized tubes described previously. As soon as possible after collection, each sample was centrifuged at 2000g for 20 minutes at 4°C. Plasma was collected and divided into 200-µL aliquots. Vascular cell adhesion molecule–1, intercellular adhesion molecule–1, and E-selectin levels were assayed by enzyme-linked immunosorbent assay technique with commercially available assay kits (BBL sVCAM-1 Assay, BBL sEselectin Assay, BBL sICAM-1 Assay; British Biotechnology Product, Abingdon, United Kingdom). Statistical analysis. All data were analyzed by analysis of variance with the Data Desk exploratory statistics package (Data Description Inc, Ithaca, NY). Factors included in the analysis of variance were the effect of gestation, the presence or absence of pre-eclampsia or gestational hypertension, and, in the second group of women, whether the source of the blood samples was the uterine or antecubital vein. Scheffé post hoc tests were used to assess the differences between subgroups of patients and an allowance was made for the effect of repeated measurements. P < .05 was taken as significant. Pearson productmoment correlation was used to assess correlations between vascular cell adhesion molecule–1 levels and markers of renal function. Defining clinical end points. Pre-eclampsia was defined as described in the classification approved by the International Society for the Study of Hypertension in Pregnancy.10 Under this classification, hypertension is defined as 1 diastolic blood pressure reading ≥110 mm Hg or 2 consecutive diastolic blood pressure readings ≥90 mm Hg ≥4 hours apart. Significant proteinuria is defined ≥300 mg total protein in a 24-hour urine collection or, if this is not available, 1+ proteinuria by dipstick on 2 consecutive occasions ≥4 hours apart. With these definitions the following classification of gestational proteinuric hypertension (pre-eclampsia) was then derived: hypertension in combination with proteinuria developing after 20 weeks’ gestation in a previously normotensive, nonproteinuric woman. This study had the approval of the Ethics Committee

466 Higgins et al

August 1998 Am J Obstet Gynecol

Table I. Basic clinical characteristics of 3 outcome groups Normotension (n = 30) Age (y) 23.7 ± 4.3 Nulliparity 100% Gestation at 39.9 ± 1.3 delivery (wk) Birth weight (g) 3559 ± 555 Induction of labor 10% Cesarean delivery 3%

Pre-eclampsia (n = 22)

Gestational hypertension (n = 13)

24.5 ± 5.1 100% 36.9 ± 3.7

28.5 ± 4.6 77% 39.9 ± 1.2

2749 ± 939 41% 55%

3853 ± 555 31% 15%

Results are expressed as mean ± SD or percentage where appropriate.

of the Rotunda Hospital, and all women in this study gave informed consent. Results Longitudinal study: Effects of pregnancy and preeclampsia on circulating vascular cell adhesion molecule–1 levels. Pre-eclampsia developed in 22 (2.2%) of the original 1001 women, and 19 of these had blood sampling repeated when the overt clinical features developed. Thirteen of these latter had a further sample taken post partum. Pre-eclampsia developed in 2 members of the pre-eclamptic group at <30 weeks’ gestation (28 weeks’ and 29 weeks’ gestation); pre-eclampsia developed after 32 weeks’ gestation in the remainder. The clinical characteristics of the groups are shown in Table I. After testing for normality, vascular cell adhesion molecule–1 levels were log transformed to allow the use of parametric analysis. Analysis of variance of vascular cell adhesion molecule–1 levels in the normotensive control group showed no significant effect related to gestation (P = .14). Post hoc tests showed significantly higher levels of vascular cell adhesion molecule–1 in the established pre-eclamptic group than in the normotensive patients at the same gestation (P = .01; Fig 1). No significant differences were recorded between the pre-eclamptic and the normotensive groups in early pregnancy or at 6 weeks post partum. Levels of vascular cell adhesion molecule–1 in the gestational hypertensive group were comparable with levels in the normotensive group but were significantly lower than that in the pre-eclamptic group (P = .02; Table II). There were weak positive correlations between vascular cell adhesion molecule–1 levels and levels of urea, creatinine, and uric acid; however, the r 2 values were only 5.4%, 13.5%, and 12.5%, respectively. Circulating vascular cell adhesion molecule–1 levels in the uteroplacental circulation compared with the peripheral circulation in normotensive and pre-eclamptic pregnancies. The mean ± SD values of maternal age, birth weight, and gestational age at delivery were, respectively, 28.5 ± 6.6 years, 3525 ± 471 g, and 38.6 ± 1.0 weeks in the

Table II. Vascular cell adhesion molecule–1 levels in normotensive and pre-eclamptic pregnancies Normotension (n = 30) 18-24 wk Median Interquartile 28-32 wk* Median Interquartile ≥33 wk Median Interquartile Post partum (6 wk) Median Interquartile

Pre-eclampsia (n = 22)

Gestational hypertension (n = 13)

540 152

560 260

— —

500 135

620 240

— —

560 330

900† 350

620 300

600 320

570 210

— —

Values are in nanograms per milliliter. *Only 2 women were in the pre-eclamptic subgroup at this gestation. †P = .01, versus normotensive group.

normotensive group and 28.2 ± 4.5 years, 2516 ± 866 g, and 35.3 ± 3.5 weeks in the pre-eclamptic group. Interim analysis showed no significant differences in intercellular adhesion molecule–1 or E-selectin levels between the pre-eclamptic and normotensive groups. Moreover, within each group there were no differences in intercellular adhesion molecule–1 and E-selectin levels between the uterine and peripheral veins (Table III). Analysis of variance of vascular cell adhesion molecule–1 results showed a significant effect related to the presence or absence of pre-eclampsia (P = .002). There was no significant effect related to the vein of origin (P = .08). Scheffé post hoc tests showed that within the normotensive group there was no difference in vascular cell adhesion molecule–1 levels between the uterine and antecubital veins (P = .55). In the pre-eclamptic group vascular cell adhesion molecule–1 levels in the antecubital vein were higher than those in the uterine vein, with a trend toward statistical significance (P = .06). Comparing the 2 groups showed that levels of vascular cell adhesion molecule–1 in the antecubital vein and the uterine vein in pre-eclampsia were significantly higher compared with those in the antecubital (P < .0001) and the uterine (P < .0001)vein in the normotensive group. Comment Our results show that circulating levels of vascular cell adhesion molecule–1 are higher in established preeclampsia than in normotensive pregnancy. This finding is in keeping with the concept of endothelial activation as central to the pathogenesis of pre-eclampsia and more specifically suggests a role for activated leukocytes in mediating this dysfunction. Preclinical increases in vascular cell adhesion molecule–1 levels have recently been re-

Higgins et al 467

Volume 179, Number 2 Am J Obstet Gynecol

Fig 1. Vascular cell adhesion molecule–1 levels in group with established pre-eclampsia, group with gestational hypertension, and normotensive control group at ≥33 weeks’ gestation. Central bars represent median values; shaded areas represent 95% confidence intervals of medians; boxes represent interquartile ranges; and whiskers represent 95th and 5th percentiles.

Table III. Vascular cell adhesion molecule–1, intercellular adhesion molecule–1,* and E-selectin* in uteroplacental and peripheral circulations Normotensive group (n = 19)

Vascular cell adhesion molecule–1 Uterine vein Antecubital vein Intercellular adhesion molecule–1* Uterine vein Antecubital vein E-selectin* Uterine vein Antecubital vein

Pre-eclamptic group (n = 20)

Median

Interquartile

Median

Interquartile

570 610

440 610

1100 1240

480† 800†

450 360

30 90

360 350

80 120

380 390

190 160

460 450

120 130

Values are in nanograms per milliliter. *Interim analysis on 12 women with pre-eclampsia and 8 normotensive women. †P < .0001, versus normotensive group.

ported.5 We did not, however, find evidence of an early preclinical increase in vascular cell adhesion molecule–1 levels in those women in whom pre-eclampsia subsequently developed. These findings are in contrast with our previous report of increased circulating levels of plasminogen activator inhibitor–1 and fibronectin (other markers of endothelial dysfunction) at 18 to 24 weeks’ gestation in women in whom pre-eclampsia subsequently developed.11 It remains to be determined whether increased vascular cell adhesion molecule–1 levels detected in our study are a marker for leukocyte-mediated tissue injury or reflect another important role for vascular cell adhesion molecule–1 in pregnancy. Greer et al2 previously reported that circulating levels

of human neutrophil elastase (a specific marker of neutrophil activation) are increased in pre-eclampsia and that this increase is confined to the maternal circulation. They suggested that activated neutrophils may cause endothelial damage. Vascular cell adhesion molecule–1 is constitutively expressed at low levels by endothelial cells, and expression can be induced by cytokines.12 It binds to the leukocyte surface molecule very late antigen-4, which is found on monocytes, lymphocytes, and eosinophils and possibly on neutrophils and thus may contribute to leukocyte-mediated tissue injury.13 Recently Gurtner et al14 showed that in mouse embryos deficient in vascular cell adhesion molecule–1 the allantois fails to fuse to the chorion, leading to abnormal placental development

468 Higgins et al

and embryonic death. This observation points to other roles for vascular cell adhesion molecule–1 in cell-cell adhesion. Our results may reflect perturbation of these additional features of vascular cell adhesion molecule–1. The mechanisms leading to the increased vascular cell adhesion molecule–1 levels in pre-eclampsia are unknown. Militating against a role for a circulating factor stimulating vascular cell adhesion molecule–1 expression is the recent report of in vitro studies by Haller et al,15 which showed that serum from patients with pre-eclampsia does not induce expression of vascular cell adhesion molecule–1 but does induce intercellular adhesion molecule–1 expression in cultured endothelial cells. However, it is not clear how closely this model reflects the in vivo mechanisms. In pre-eclampsia, cytokine-induced expression is a potential explanation. Tumor necrosis factor–α is known to increase vascular cell adhesion molecule–1 expression.16 Vince et al17 demonstrated increased circulating levels of both tumor necrosis factor–α and tumor necrosis factor–α receptors in pre-eclampsia. Arguing against a role for tumor necrosis factor–α, however, this cytokine increases vascular cell adhesion molecule–1, intercellular adhesion molecule–1, and E-selectin in cultured endothelial cells. Thus selective expression of vascular cell adhesion molecule–1 in pre-eclampsia is more difficult to explain. Other cytokines, such as interleukin 4, may be involved.16 Alternatively, oxidant stress may be the stimulus. By analogy, early atherosclerotic lesions are characterized by increased oxidative stress, which is evidenced by the increased production of reactive oxygen species and increased expression of vascular cell adhesion molecule–1 in atherosclerotic plaques.18 These oxygen species convert low-density lipoprotein to the endothelial cytotoxic oxidized low-density lipoprotein. Oxidative stress and the selective expression of vascular cell adhesion molecule–1 may be causally linked. Marui et al19 showed that cytokineactivated vascular cell adhesion molecule–1 gene expression in cultured human umbilical vein endothelial cells may be almost completely inhibited by the addition of antioxidants. In contrast, addition of antioxidants does not suppress the cytokine-induced intercellular adhesion molecule–1 gene expression. Thus there is evidence of an antioxidant-sensitive transcriptional regulatory mechanism for vascular cell adhesion molecule–1 gene expression. Pre-eclampsia is now known to be associated with oxidative stress and increased levels of lipid peroxide combined with decreased levels of endogenous and exogenous antioxidants have been shown in pre-eclampsia.20, 21 It has been recently suggested that hypertension may in itself induce oxidative stress on the endothelium.22 Our results show no evidence of increased vascular cell adhesion molecule–1 levels in patients with pure gestational hypertension. This finding underlines the importance of distinguishing gestational hypertension from

August 1998 Am J Obstet Gynecol

pre-eclampsia. It is in keeping with the landmark studies of Chesley et al,23 which showed that these conditions behave very differently on long-term follow-up. It also concurs with work from our own unit, which has shown that women with gestational hypertension and pre-eclampsia have different hemodynamic characteristics as early as 18 to 24 weeks’ gestation.9 In considering whether raised vascular cell adhesion molecule–1 levels simply reflect decreased renal clearance, we found a statistically significant but weak correlation between vascular cell adhesion molecule–1 levels and levels of urea, creatinine, and uric acid. It is therefore unlikely that decreased renal clearance is a cause of the increased circulating levels of vascular cell adhesion molecule–1. In addition, we can state that increased vascular cell adhesion molecule–1 levels in the group with established pre-eclampsia are not caused by a gestational effect because we found no such change in vascular cell adhesion molecule–1 levels throughout normal pregnancy and the puerperium. The pathologic placental findings in pre-eclampsia of microthrombi, excessive fibrin deposition, and foamy macrophages have been described as “acute atherosis.” Ultrastructural studies by Shanklin and Sibai24 of the placental bed and uterine boundary arteries in pre-eclampsia revealed swollen and eroded endothelium. Our study also addresses the question of whether the uteroplacental circulation is the source of increased vascular cell adhesion molecule–1 levels. We showed significantly higher levels of vascular cell adhesion molecule–1 in the uteroplacental and peripheral circulations of women with preeclampsia undergoing cesarean delivery than in normotensive control subjects. There was no evidence of a concentration gradient between the uteroplacental and the peripheral circulations in either group. These findings would be in keeping with the work of Lyall et al,25 who showed no vascular cell adhesion molecule–1 expression in placental tissue from normotensive and preeclamptic women in late pregnancy. One possible difficulty in interpreting our data was the concern that the results would be confounded by the effect of surgery itself on vascular cell adhesion molecule–1 levels. It is therefore reassuring that levels of vascular cell adhesion molecule–1 measured in the sample taken from the antecubital vein in the control group at the time of cesarean delivery were similar to those found in the samples taken from the antecubital vein in our longitudinal control group. Our results show that the endothelium is an important target of injury in pre-eclampsia. The development of clinical pre-eclampsia was associated with a striking increase in vascular cell adhesion molecule–1 levels, which resolved after delivery. This difference was not observed with intercellular adhesion molecule–1 or E-selectin, suggesting selective disturbance of vascular cell adhesion molecule–1 homeostasis. Vascular cell adhesion mole-

Higgins et al 469

Volume 179, Number 2 Am J Obstet Gynecol

cule–1 levels were not elevated in gestational hypertension, suggesting that vascular cell adhesion molecule–1 dysregulation was specific for pre-eclampsia and not a generalized response to hemodynamic stress. The source of vascular cell adhesion molecule–1 appears to be the peripheral rather than the uteroplacental circulation. Importantly, there was at best only a weak correlation between circulating vascular cell adhesion molecule–1 and either urea, creatinine, or urate, suggesting increased vascular cell adhesion molecule–1 release into the circulation rather than decreased renal clearance. In aggregate these studies provide evidence for differential activation of endothelial responses in pre-eclampsia and suggest that the role of vascular cell adhesion molecule–1 in the pathophysiology of this important condition is worthy of further investigation. We gratefully acknowledge the contribution of research midwives Anna O’Sullivan and Audrey McMahon to the completion of this study.

10.

11.

12. 13.

14.

15.

16.

17.

REFERENCES

1. Roberts JM, Redman CW. Pre-eclampsia: more than pregnancyinduced hypertension. Lancet 1993;341:1447-51. 2. Greer IA, Dawes J, Johnson TA, Calder AA. Neutrophil activation is confined to the maternal circulation in pregnancy-induced by hypertension. Obstet Gynaecol 1991;78:28-32. 3. Picker LJ. Mechanisms of lymphocyte homing. Curr Opin Immunol 1992;4:277-86. 4. Gearing AJ, Hemingway I, Pigott R, Hughes J, Rees AJ, Cashman SJ. Soluble forms of adhesion molecules, E-selectin, ICAM-1 and VCAM-1: pathological significance. Ann N Y Acad Sci 1992;667:324-31. 5. Krauss T, Kuhn W, Lakoma C, Augustin HG. Circulating endothelial cell adhesion molecules as diagnostic markers for the early identification of pregnant women at risk for the development of preeclampsia. Am J Obstet Gynecol 1997;177:443-9. 6. Austgulen R, Lien E, Vince G, Redman CWG. Increased maternal plasma levels of soluble adhesion molecules (ICAM-1, VCAM-1, E-selectin) in preeclampsia. Eur J Obstet Gynecol Reprod Biol 1997;71:53-8. 7. Lyall F, Greer IA, Boswell F, Macara LM, Walker JJ, Kingdom JC. The cell adhesion molecule, VCAM-1, is selectively elevated in serum in pre-eclampsia: does this indicate the mechanism of leucocyte activation? Br J Obstet Gynaecol 1994;101:485-7. 8. Djurovic S, Schjetlein R, Wisloff F, Haugen G, Berg K. Increased levels of intercellular adhesion molecules and vascular cell adhesion molecules in pre-eclampsia. Br J Obstet Gynaecol 1997;104:466-70. 9. Higgins JR, Walshe JJ, Halligan A, O’Brien E, Conroy R, Darling MR. Can 24-hour ambulatory blood pressure predict the devel-

18.

19.

20.

21.

22.

23.

24.

25.

opment of hypertension in primigravidae? Br J Obstet Gynaecol 1997;104:356-62. Davey DA, MacGilllivray I. The classification and definition of the hypertensive disorders of pregnancy. Am J Obstet Gynecol 1988;158:892-8. Halligan A, Bonnar J, Sheppard B, Darling M, Walshe J. Haemostatic, fibrinolytic and endothelial variables in normal pregnancies and pre-eclampsia. Br J Obstet Gynaecol 1994;101:488-92. Brady HR. Leukocyte adhesion molecules and kidney diseases. Kidney Int 1994;45:1285-300. Elices MJ, Osborn L, Takada Y, Crouse C, Luhowskyj S, Hemler ME, et al. VCAM-1 on activated endothelium interacts with the leukocyte integrin VLA-4 at a site distinct from the VLA-4/fibronectin binding site. Cell 1990;60:577-82. Gurtner G, Davis V, Li H, McCoy N, Sharpe A, Cybulsky MI. Targeted disruption of the murine VCAM-1 gene: essential role of VCAM-1 in chorioallantoic fusion and placentation. Genes Dev 1995;9:1-14. Haller H, Ziegler EM, Homuth V, Drab M, Eichorn J, Nagy Z, et al. Endothelial adhesion molecules and leukocyte integrins in preeclamptic patients. Hypertension 1997;29:291-6. Thornhill MH, Wellicome SM, Mahiouz DL, Lanchbury JL, Kyan-Aung U, Haskard DO, et al. Tumor necrosis factor combines with IL-4 or IFN-γ to selectively enhance endothelial cell adhesiveness for T cells. J Immunol 1991;146:592-8. Vince GS, Starkey PM, Austgulen R, Kwiatkowski D, Redman CW. Interleukin-6, tumour necrosis factor and soluble tumour necrosis factor receptors in women with pre-eclampsia. Br J Obstet Gynaecol 1995;102:20-5. Cybulsky MI, Gimbrone MA. Endothelial expression of a mononuclear leukocyte adhesion molecule during atherogenesis. Science 1991;251:788-91. Marui N, Offermann MK, Swerlick R, Kunsch C, Rosen CA, Ahmad M, et al. Vascular cell adhesion–1 (VCAM-1) gene transcription and expression are regulated through an antioxidantsensitive mechanism in human vascular endothelial cells. J Clin Invest 1993;95:1866-74. Wisdom SJ, Wilson R, McKillop JH, Walker JJ. Antioxidant systems in normal pregnancy and in pregnancy induced hypertension. Am J Obstet Gynecol 1991;165:1701-4. Wang Y, Walsh S, Guo J, Zhang J. The imbalance between thromboxane and prostacyclin in preeclampsia is associated with an imbalance between lipid peroxides and vitamin E in maternal blood. Am J Obstet Gynecol 1991;165:1695-700. Alexander RW. Oxidative stress and the mediation of arterial inflammatory response: a new perspective. Hypertension 1995;25: 155-61. Chesley LC, Annitto JE, Cosgrove RA. The remote prognosis of eclamptic women: sixth periodic report. Am J Obstet Gynecol 1976;124:446-59. Shanklin DR, Sibai BM. Ultrastructural aspects of preeclampsia: I. Placental bed and uterine boundary vessels. Am J Obstet Gynecol 1989;161:735-40. Lyall F, Greer IA, Boswell F, Young AA, Macara LM, Jeffers MD. Expression of cell adhesion molecules in placentae from pregnancies complicated by pre-eclampsia and intrauterine growth retardation. Placenta 1995;16:579-87.