An alternative approach to the pathogenesis of preeclampsia is that of ischemia-reperfusion. Over recent years there has been an explosion of interest in this field in relation to cerebrovascular and cardiovascular insults (Carden and Granger, 2000; Collard and Gelman, 2001; Grace and Mathie, 1999), and there is increasing evidence that much of the pathophysiology involved can also be applied to the placenta. Underlying ische-mia-reperfusion injury must be intermittent perfusion, and in the case of the placenta this can occur through two mechanisms: intrinsic vasoconstriction of the spiral arteries, and external compression of the arcuate and radial arteries during uterine contractions.
The spiral and radial arteries have a thick muscular wall, and vasoconstriction of these vessels plays a key role in preventing excessive blood loss during menstruation. Because the proximal unconverted segment of the spiral artery retains its smooth muscle coat it remains vaso-reactive even in normal pregnancies. This has been clearly demonstrated in the rhesus monkey, in which the intravenous administration of epineph-rine or norepinephrine causes a severe reduction in placental perfusion, to the point that high doses can result in fetal asphyxia and death (Adamsons and Myers, 1975; Adamsons et al., 1971). Evidence that the unconverted segment of the vessel is indeed flow-limiting is provided by the fact that despite the systemic hypertension induced the pressure in the dilated mouth of the spiral artery is decreased after administration. It has been suggested that a similar mechanism underlies the high rate of fetal loss during the second and third trimesters in women suffering from pheochromo-cytoma (Adamsons and Myers, 1975). Spontaneous changes in maternal catecholamine release, for example in response to stress, may therefore also influence the caliber of these vessels and hence perfusion of the intervillous space. Epidemiologically, there is an association between maternal anxiety and low birthweight, and the finding that women with high anxiety scores display increased uterine artery resistance may explain the link (Teixeira et al., 1999).
More direct evidence that constriction of the proximal sections of the arteries may occur in vivo is again largely based on the rhesus monkey, in which cineradioangiography has revealed that flow from many spiral arteries into the intervillous space is intermittent (Martin et al., 1964; Ramsey and Donner, 1980, 1988). When performing repeated injections of contrast medium into the aorta at intervals of 5—30 min, Martin et al. observed that whilst discharge occurred from some spiral arteries following each injection, discharge from other arteries occurred after some injections but not others. The proportion of the arteries demonstrating such intermittent flow ranged from 1 in 7 in one female to 14 of 28 in another, and a larger number of vessels showed variations in the size of the intervillous spurts between injections. There was no association with gestational age, although the sample size was small, and the arteries were located both centrally and peripherally within the placenta. Because the injections were timed to coincide with periods of uterine relaxation, Martin et al. concluded that the effect was due to vasoconstriction within individual arteries rather than external compression caused by myometrial contractions.
There is no doubt from observations on the rhesus monkey that uterine contractions can have a profound impact on intervillous blood flow (Ramsey and Donner, 1980). During contractions the pressure rises in the different fluid compartments of the feto-placental unit and the normal differential that drives maternal blood through the intervillous space will be disrupted. In addition, external compression of the artery during the contraction may occlude the lumen. By placing a cannula in the amniotic cavity Ramsey et al. were able to measure the pressure within the cavity and so time injections of contrast medium with relation to the cycle of uterine contractions (Ramsey et al., 1963). They observed that inflow of maternal blood into the intervillous space is either abolished or greatly reduced during contractions depending on the strength of the latter. Thus in one animal at 64 days of gestation entry was observed from 18 spiral arteries following injection during the relaxation phase. By contrast, entry was observed only from two arteries after injection during a strong contraction, and from six during a milder contraction. The strength of the contractions varied with gestational age, being strongest in early and late pregnancy when they either abolished or severely reduced inflow, and weaker during mid-pregnancy when they had little effect.
Such detailed experiments are not possible for the human, but there is limited evidence from the literature that intermittent perfusion of the inter-villous space does occur. Separating the phenomena of individual arterial constriction and external compression during uterine contractions is difficult, as serial imaging cannot be performed. None the less, evidence for the latter was provided by Borell and colleagues. These authors performed angiography at 17—20 weeks of gestation in patients prior to termination of pregnancy, and induced uterine contractions with oxytocin or intra-amniotic injections of saline (Borell et al., 1965a). Contrast medium introduced into the aorta entered the intervillous space within 4—5 s in the relaxed state, whereas during contractions entry was only observed in around 50% of cases. Furthermore, the number of entry sites was reduced during contractions, as was the volume of contrast medium discharged. The same group performed similar investigations on three women carrying severely malformed fetuses at or near term (Borell et al., 1965b). They observed both local and generalized narrowing of large and small arterial branches within the uterine wall during contractions. No definite changes in the diameter of the spiral arteries could be detected, a feature that the authors attributed to the fact that these lie predominantly within the endometrium where local external pressure does not occur.
These data are therefore in accord with those from the rhesus monkey. More recently, magnetic resonance imaging techniques have been applied to the human, both in normal and growth restricted pregnancies (Francis et al., 1998). As studies have only been performed at a single time point they cannot provide information on the constancy of intervillous blood flow. None the less, they have revealed significant regional differences in flow in both groups, indicating that intervillous flow is not homogeneous throughout the organ.
Incomplete conversion of the spiral arteries will most likely exacerbate this situation, for the retention of a greater length of vasoreactive muscular spiral artery must predispose to more spasmodic intervillous flow. We suggest that this effect will have a more profound impact in the pathogenesis of pre-eclampsia than the diminished caliber of the terminal funnel.
Was this article helpful?