Introduction

Although the gross anatomy of the maternal blood supply to and drainage from the intervillous space was well documented by William Hunter [1] in 1774 a considerable degree of confusion persisted, and in particular the understanding of the anatomical structure of the ‘curling’ arteries remained incomplete and often not based on data. Therefore as an introduction to the chapters on placental bed vascular disorders the early literature on the maternal blood supply to the placenta is briefly reviewed.

The term ‘placental bed’ was introduced 50 years ago by Dixon and Robertson [2] and can be grossly described as that part of the decidua and adjoining myometrium which underlies the placenta and whose primary function is the maintenance of an adequate blood supply to the intervillous space of the placenta. Certainly, there is no sharp anatomical demarcation line between the placental bed and the surrounding structures, but, as this part of the uterine wall has its own particular functional and pathological aspects, it has proven to be a most useful term for describing the maternal part of the placenta in contrast to the fetal portion.

The uteroplacental arteries

Anatomically the uteroplacental arteries can be defined as the radial and spiral arteries which link the arcuate arteries in the outer third of the myometrium to the intervillous space of the placenta. Before reaching the myometrio-decidual junction, the radial arteries usually split into two or three spiral arteries. When they enter the endometrium the spiral arteries are separated from each other by a 1–6 mm gap [3]. Small arteries, the so-called basal arterioles, branch off from the proximal part of the spiral arteries and vascularize the myometrio-decidual junction and the basal layer of the decidua. They are considered to be less responsive, if at all, to cyclic maternal hormones [4].

Two comments are appropriate here. First, some confusion existed as to whether the spiral arteries of the placental bed should be called ‘arteries’, which was commonly used in German literature, or ‘arterioles’, which was more common in the English literature. In view of the size of the spiral vessels, which communicate with the intervillous space, the terminology of ‘spiral arteries’ was adopted for these vessels in order to distinguish them from the ‘spiral arterioles’ of the decidua vera. A second comment relates to the spiral course as described by Kölliker [5]. Because during pregnancy these arteries increase in length as well as in size, Bloch [6] suggested that in the human the terminal part of the spiral artery is no longer spiral or cork-screw, but has a more undulating course as was demonstrated in the Rhesus monkey (Macaca mulatta) by Ramsey [7].

The origin of placental septa and the orifices of spiral arteries have been the subject of great controversy. Bumm [8,9] pointed out that the arteries are mainly lying in the decidual projections and septa, and eject their blood from the side of the cotyledon into the intervillous space (Fig. 2.1). Bumm’s statement has been quoted as implying that the arteries open in the intervillous space near the chorionic plate, while Bumm obviously regarded the subchorionic blood as somewhat venous in nature. Wieloch [10] and Stieve [11] have corrected Bumm’s observation in that they specified that the spiral arteries mainly open at the base of the septa. Boyd and Hamilton [4] confirmed that the septa are of dual maternal and trophoblastic origin and that arterial orifices are scattered more or less at random over the basal plate. The orifices of several arteries may be grouped closely together and, in individual vessels, are usually at their terminal portions. Spiral arteries may have initially multiple openings. Such multiple openings can later become separated by the straightening out and dilatation of the artery and the unwinding of the coils during placental growth. When multiple openings are present the segment of the artery between successive ones may show obliteration of the lumen.

Fig. 2.1 Diagrammatic representation of the course of the maternal circulation through the intervillous space of the placenta. After Bumm [9].

Attempts to count the number of spiral arteries communicating with the intervillous space have been made by several investigators. Klein [12] counted in one separated placenta 15 maternal cotyledons with 87 arteries and 39 veins and in a second 10 cotyledons with 45 arteries and 27 veins. Spanner [13] working with a corrosion preparation of about 6 months’ gestation found 94 arteries communicating with the intervillous space. Boyd [3] made calculations of total numbers based on sample counts of openings of the spiral arteries in the basal plate in three placentae of the third and fourth months of pregnancy. The calculations for full-term placentae varied between 180 and 320 openings but all three counts can only be considered as first appreciations. On the other hand, Ramsey [7] showed by serial sections in the Rhesus monkey the uneven distribution of arterial communications with the intervillous space and suggested that partial counts may have introduced errors in the calculations. In an anatomical reconstruction of two-fifths of the maternal side of a placenta in situ at term Brosens and Dixon [14] confirmed the irregular arrangement of septa and arterial and venous openings. All arteries opened into the intervillous space by a solitary orifice. They found in a normal placenta 45 openings for a surface area of 32 cm² [15] and in a uterus with placenta in situ from a woman with severe preeclampsia 10 spiral arteries for a surface area of 7 cm² [16], which in both cases amounts to one spiral artery for every 0.7 cm² of placental bed.

A bird’s-eye view of the three-dimensional basal plate shows septa of various sizes with the majority of arterial orifices at the base of a septum. Septa are likely to represent uplifted basal plate reflecting differences in depth of decidual trophoblast invasion and resulting in a conchiform base for the anchoring of a fetal cotyledon. Arterioles high up in the septa and without an orifice into the intervillous space are likely to be basal arterioles.

The anatomy of the venous drainage has also been the subject of much discussion. Kölliker [5] stressed in 1879 the role of the marginal sinus, partly lying in the placenta and partly in the decidua vera. Spanner [13] revived this theory in 1935, however without quoting Kölliker. The anatomical work by many authors has subsequently shown that venous drainage occurs all over the basal plate. The veins fuse beneath the basal plate to form the so-called ‘venous lakes’ [7]. The term ‘sinusoid’ has been applied to these vessels, but has caused much confusion in the literature as the term has been used for the intervillous space and the spiral arteries.

The question of arteriovenous anastomoses in the decidua arose when Hertig and Rock [17] described extensive anastomoses in the decidua. Bartelmez [18], however, after re-examination of the original histological sections of Hertig and Rock [17] cast doubt on the drawings published by these authors in 1941 and the existence of such shunts was later disproved. Recently, Schaaps and collaborators [19] used three-dimensional sonography and anatomical reconstruction to investigate the placental bed vasculature and demonstrated an extensive vascular anastomotic network in the myometrium underlying the placenta. No such network was seen outside the placental bed. It can be speculated that the subendometrial network is formed by the hypertrophied basal arterioles and veins in the placental bed.

Pregnancy changes, intraluminal cells and giant cells

The morphological changes of the uteroplacental arteries were extensively studied, mainly by German investigators, around the turn of the last century and particularly in the context of the mechanism preventing the uterus from bleeding during the postpartum period.

Almost all authors before 1925 agreed that thickening of the intima occurs in myometrial arteries of the uterus as a result of gestation. Wermbter [20] in an extensive study showed that this change is not specific for pregnancy, but is also related to some degree with parity. The importance attached by these authors to intimal thickening was that under the influence of contractions the vessel becomes occluded and that the projections caused by the intimal proliferation could act as supports for the formation of thrombi causing primary occlusion of the vessel in the postpartum period. The large myometrial arteries of the multigravida are characterized by an abundance of elastic and collagenous tissue in the adventitia, although the amount of increase in elastic tissue does not necessarily correlate with the number of gestations.

While most authors seemed to be agreed on the changes in the myometrial arteries much confusion and discussion existed with respect to vessel changes in the placental bed. Friedländer [21] described in 1870 an outstanding vessel change in the placental bed, which was wrongly described by Leopold [22] as ‘die Spontane Venenthrombose’. Friedänder’s description is as follows:

In 1904 Schickele [23] drew attention to the fact that the vessel changes described by Friedländer [21] and Leopold [22] occurred mainly in arteries and only occasionally in veins. However, they were incorrect in thinking that the cells in the arterial lumen were most marked in late pregnancy as their description included two different changes which, although related to each other, appear in the spiral arteries at a different time during the course of pregnancy. A confusing terminology has been used to describe the changes which occur in the wall of the spiral arteries communicating with the intervillous space, such as ‘physiologische regressive Metamorphose’, ‘hyalin Rohr mit grössen Zellen’, fibrinoid and hyaline structures of bizarre outline in collapsed vessels, and diffuse thickening of the entire wall.

The intrusive cells in the lumen as described by Friedländer [21] were intensively studied by Boyd and Hamilton [24] and Hamilton and Boyd [25,26] using their large collection of uterine specimens with the placenta in situ. They demonstrated the continuity of these cells with the cytotrophoblastic cells of the basal plate of the placenta. The intraluminal cells first appear in the arteries when the latter are being tapped by the invading trophoblast; the maternal blood then reaches the intervillous space by percolating through the gaps between the intraluminal cells. They decided that the most acceptable explanation was that these cells were derived from the cytotrophoblastic shell and migrated antidromically along the vessel lumen. The intraluminal cells can pass several centimeters along a spiral artery and, indeed, may be found in its myometrial segment. Such plugging by intraluminal cells was illustrated in a myometrial artery from a pregnant uterus with a fetus of 118 mm CR length [4]. The plugs were present in all the spiral arteries of the basal plate during the middle 3 months of pregnancy, although their numbers varied, and they disappeared altogether in the last months. They were never observed in the veins. Boyd and Hamilton [4] speculated that the intravascular plugs damped down the arterial pressure in arteries that had already lost their contractility.

Kölliker [5] was the first to describe in 1879 the giant cells (‘Riesenzellen’) in the placental bed and indicated that these cells are restricted to the decidua basalis. Opinions diverged on the origin of these cells. The fetal origin was demonstrated by Hamilton and Boyd [26] when they examined uteri with placenta in situ at closely related time intervals during pregnancy and observed continuity in the outgrowth of fetal syncytial elements into the maternal tissue. Suggested functions of the giant cells were the production of enzymes, possibly to ‘soften up’ the maternal tissue, and the elaboration of hormones. Hamilton and Boyd had the impression that there was no marked effect, cytolytic or otherwise, of the giant cells on the maternal tissue. These cells seemed to push aside the maternal cells and to dissolve the surrounding reticulin and collagen, but there was no apparent destructive effect on the adjacent maternal cells. The possibility of hormone production by giant cells was suggested by their histological and histochemical appearance.

Functional aspects

In the early 1950s the hemodynamic aspects of the maternal circulation of the placenta were investigated using different new functional techniques such as the determination of the ²⁴Na clearance time in the intervillous space [27], cineradiographic visualization of the uteroplacental circulation in the monkey [28,29], and determination of the pressure in the intervillous space [30].

Measurements of the amount of maternal blood flowing through the uterus and the intervillous space made by Browne and Veall [27] and Assali and co-workers [31] showed that maternal blood flows through the uterus during the third trimester at a rate of approximately 750 ml/min, and 600 ml/min through the placenta. Browne and Veall [27] found a slight but progressive slowing of the flow in late pregnancy up to term. However, in the presence of maternal hypertension a considerable decrease of flow was found and the extent of change appeared to be related to the severity of maternal hypertension.

Pathology of uteroplacental arteries

Vascular lesions of the uteroplacental arteries have been described since the beginning of the last century. Seitz [32] described in 1903 the intact uteri with placentae from two eclamptic patients with abruptio placentae and noted a proliferative and degenerative lesion in the spiral arteries, with narrowing and even occlusion of the vascular lumen. He found occluded arteries underlying an infarcted area of the in situ placenta, and related the vascular and decidual degeneration to the toxemic state. In later literature this excellent report on the uteroplacental pathology in eclampsia has been completely ignored, probably because at that time most authors were mainly interested in the presence of inflammatory cells in the decidua as a possible cause of eclampsia.

The delivered placenta and fetal membranes were for many years the commonest method of obtaining material for the study of spiral artery pathology, and there were large discrepancies between the findings in this material. In preeclampsia lesions such as acute degenerative arteriolitis [33], acute atherosis [34], and arteriosclerosis [35] were described.

Dixon and Robertson [2] introduced 50 years ago at the University of Jamaica the technique of placental bed biopsy at the time of cesarean section, while the Leuven group [36,37] obtained biopsies after vaginal delivery using sharpened ovum forceps. Both groups described hypertensive changes that showed the characteristic features of vessels exposed to systemic hypertension, i.e. hyalinization of true arterioles and intimal hyperplasia with medial degeneration and proliferative fibrosis of small arteries.

Physiological changes of placental bed spiral arteries

The method of placental bed biopsy produced useful material, but nevertheless was criticized by Hamilton and Boyd (personal communication). They strongly recommended the examination of intact uteri with the placenta in situ for the simple reason that the placental bed is such a battlefield that fetal and maternal tissues are hard to distinguish on biopsy material and maternal vessels are disrupted after placental separation. In 1958, independent from the British group in Jamaica, the Department of Obstetrics and Gynaecology of the Catholic University of Leuven had also started to collect placental bed biopsies, and in 1963 they began to collect uteri with the placenta in situ [37,38]. The hysterectomy specimens were obtained from women under normal and abnormal conditions whereas today tubal sterilization would have been performed at the time of cesarean section. The technique for keeping the placenta in situ at the time of cesarean hysterectomy was rather heroic. Immediately after delivery of the baby the uterine cavity was tightly packed with towels in order to reduce uterine retraction and prevent the placenta from separating from the wall. The large uterine specimens with placenta in situ were examined by semiserial sections to trace the course of spiral arteries from the basal plate to deep into the myometrium. As a result Brosens, Robertson and Dixon [39] described in 1967 the structural alterations in the uteroplacental arteries as part of the physiological response to the pregnancy and introduced for these vascular adaptations the term ‘physiological changes’ (Fig. 2.2). In 1972 the same authors [40] published the observation that preeclampsia is associated with defective physiological changes of the uteroplacental arteries in the junctional zone myometrium.

Fig. 2.2 Diagram of the maternal blood supply to the placental bed and intervillous space of the placenta showing physiological changes of the spiral arteries in the basal plate, decidua and junctional zone myometrium. After Brosens et al. [39].

In subsequent studies the remodeling of the spiral arteries was investigated during the early stages of pregnancy. While abortion for medical reasons was allowed in the UK, it was not uncommon for older women to have a hysterectomy. When Geoffrey Dixon moved to the Academic Department of Obstetrics and Gynaecology of the University of Bristol in the 1970s he started to collect uteri with the fetus and placenta in situ from terminations of pregnancy by hysterectomy. The Bristol collection of uteri with placenta in situ was the starting point for the study of the development of uteroplacental arteries by Pijnenborg and colleagues [41].

Conclusions

The history outlined above illustrates the vascular complexity of deep placentation in humans. The spiral artery anatomy as well as the vascular pathology were only revealed after studying uteri with in situ placentae. There is no doubt that the main issue has been the recognition of the structural adaptation of the spiral arteries in the placental bed and the association of defective deep placentation with clinical conditions such as preeclampsia.

The main challenge today is to understand the mechanisms of the vascular adaptations and the role of the trophoblast and the maternal tissues in the interactions that can lead to a spectrum of obstetrical disorders.