Shifting Concepts of the Fetal–Maternal Interface: A Historical Perspective
Article Outline
- Abstract
- 1. Introduction
- 2. Carpenter's confusion: the delineation of the intervillous space
- 3. Huxley's intuition: the importance of the decidua
- 4. Duval's inspiration: the invasion of maternal tissue
- 5. Hubrecht's assertion: the naming of the trophoblast
- 6. Nitabuch's allegation: the fibrinoid as a barrier to trophoblast invasion
- 7. Conclusion
- 8. Conflict of interest
- Acknowledgement
- References
- Copyright
Abstract
The first microscopic images of the human placenta, obtained in the 1830s, revealed the presence of an epithelial lining separating fetal capillaries from maternal blood, which was in later years successively interpreted as maternal endothelial, decidual and finally as “trophoblastic”. With this new term, introduced by Hubrecht in 1889, its embryonic/fetal origin was recognized as well as its role in nutrient uptake from maternal blood. Thomas Huxley considered the presence of a decidua as an important feature for mammalian classification, but still mixed up maternal and trophoblastic tissue. Mathias Duval recognized invasive activities by trophoblast in rodents, but over-interpreted the arterial invasion observed in rats. In the human, unusual endovascular cells were first described by Carl Friedländer, but their trophoblastic nature was only recognized in the early 20th century. Nitabuch's description of a continuous fibrinoid layer underneath the basal plate led to the erroneous concept of a borderline separating the trophoblast-invaded upper decidua from the deeper non-invaded uterine tissue. This concept – based on the study of one pregnant uterus – has been made obsolete by later studies of trophoblast invasion. Many erroneous interpretations of placental histology in the past were logical in the context of then current knowledge. A better understanding depended on improved technology which allowed tracing of histological continuity of structural features in space and time. Although identification of cell types increasingly relies on molecular markers, classical histological principles should still be applied in conjunction with newer techniques in order to arrive at a broad understanding of placental development. Understanding past errors in interpreting placental histology should guard us against overconfidence in so-called breakthrough discoveries.
Keywords: Placental histology, Trophoblast, History, Fetal–maternal interface
1. Introduction
A milestone in the history of placentology was undoubtedly the definite proof, delivered by John Hunter (1728–1793), that maternal and fetal circulations in the placenta are not continuous but remain separate [1]. Previously, his brother William (1718–1783) had described the “convoluted” (spiral) arteries of the uterus which deliver maternal blood to the placenta, as confirmed by coloured wax injection experiments. During the same time period, Antoine-Laurent de Lavoisier (1743–1794) discovered the element oxygen and pointed out similarities between the processes of combustion and respiration. This new physiological concept was actively promoted in England by Erasmus Darwin (1731–1802), who argued in his major work “Zoonomia” (1794–1796) that the concept was also applicable to placental function. Indeed, oxygen uptake could be visualized by the colour change of the blood in lungs, gills of fish, and – as a striking analogue of the latter – also in the placenta. This new insight, however, tended to eclipse temporarily the previously assumed role of delivering nutrients to the developing fetus [2], [3].
Further advances in the understanding of placental function necessitated a detailed knowledge of its microscopic structure. In the early 19th century microscopes were technically still inadequate. Probably the first acceptable microscopic images of the human placenta were obtained in 1832 by Ernst Heinrich Weber. After teasing out placental tissue with fine needles he noticed under the microscope that fetal capillaries are enclosed within villous structures covered by a pellucid membrane of undetermined origin. Unfortunately Weber never reported his observations in scientific papers of his own, but generously passed his findings to others who included them in anatomical and physiological textbooks [4]. Until then, anatomists had considered that fetal blood vessels might run free within the maternal blood spaces, and the presence of a membrane surrounding loops of fetal capillaries was therefore a new finding. The cellular nature of this limiting membrane was not immediately clear, but it was suggested that some kind of “chorionic epithelium” separated the maternal blood from the fetal capillaries. Previously, substances in the maternal blood were thought to be absorbed by the fetus via tiny fetal vessel openings into maternal blood spaces of the placenta [3], but the existence of an extra-intervening membrane between the two circulations necessarily implied that this unidentified layer had to serve as a passage-way for maternal to fetal transport.
2. Carpenter's confusion: the delineation of the intervillous space
Early in the 19th century it might have been difficult to conceive how the maternal blood could be expelled from opened spiral arteries into an intervillous space. At first sight such a vascular arrangement might have been considered a violation of the principle of the closed vascular system existing in vertebrates. Early placental histologists, however, arrived at a plausible solution for this problem. In a widely used textbook of comparative physiology published in 1854, William Carpenter (1813–1885) described the human placenta as comprising fetal “vascular tufts … [which] dip-down, as it were, into a cavity, whose inner wall is formed by an extension of the lining membrane of the large veins or sinuses of the uterus” [5]. By considering the intervillous space as dilated, endothelial-lined, uterine sinuses, the principle of a closed vascular system was saved. Carpenter added an illustration showing histological details, noting that “the extremity of every loop of foetal vessels forming a placental villus is invested first by a foetal, and then by a maternal, layer of cells” (Fig. 1). The bilayered disposition was particularly clear when clefts appeared between the two layers, presumably as artefacts of tissue processing.
Fig. 1
Structure of human placental villus according to Carpenter [5]: “Extremity of a placental villus: a, external membrane of the villus, continuous with the lining membrane of the vascular system of the mother; b, external cells of the villus, belonging to the placental decidua; c, c, germinal centres of the external cells; d, the space between the maternal and foetal portions of the villus; e, the internal membrane of the villus, continuous with the external membrane of the chorion; f, the internal cells of the villus, belonging to the chorion; g, the loop of umbilical vessels” (original legend). In reality layer b represents the syncytiotrophoblast which is overlying the cytotrophoblast layer f. The space d probably is a processing artefact.
Oddly enough, this early reference never turns up in the placenta literature. Following our present-day terminology, Carpenter's inner (“fetal”) layer of cells might correspond to the cytotrophoblast layer (if it is a first trimester specimen), while his outer (“maternal”) layer represents the syncytiotrophoblast. The discovery of a villous epithelial layer (the cytotrophoblast layer) is usually ascribed to Theodor Langhans (1839–1915), who published his original findings in 1877. Langhans himself was uncertain about the nature and origin of this and the overlying syncytial layer, changing his mind repeatedly in his later publications. The rather confusing literature about this question was reviewed by Otto Grosser (1873–1951) in 1927 [6]. He listed more than 10 different opinions about the origin of the covering layers of the villi, including the rather fanciful idea that they might be derived from the corona cells which surround the ovulated oocyte.
In 1877 William Turner (1832–1916) [7] had compiled the current understanding of comparative placentation, pointing out that structural variations in placentas of different species reflect the degree to which originally separable maternal and fetal parts had become intertwined. He thought that the human placenta had acquired the highest level of specialization, since “the maternal vessels in the substance of the placenta have not only lost their tubular cylindrical form, but have become expanded into a freely communicating series of irregular, cavernous, intra-placental sinuses, so that their derivation from the vessels of the non-gravid mucosa is, at the first sight, difficult to recognize”. His interpretation of placental histology was therefore similar to Carpenter's: he considered that maternal endothelium covers the villous surfaces while the fetal (chorionic) epithelial layer is no longer recognizable. Moreover, some of the villi “are anchored to the layer of modified uterine mucosa, named decidua serotina, which walls in the placenta on the uterine aspect, by slender trabecular prolongations of the serotina.” [8] (Fig. 2). In other words, what we know now as the cytotrophoblastic cell columns of anchoring villi were interpreted as foci of decidual cells growing out into the placental tissue. This concept that the intervillous space was an intrinsic part of the uterus implied that there was only one maternal–fetal interface. The gradually emerging fact that the intervillous space is external to the uterine tissue necessitated consideration of a quite different interface between mother and fetus: the region of attachment of the placenta to the uterus, a boundary zone where maternal and fetal tissues are in intimate juxtaposition.
Fig. 2
Structure of an anchoring villus according to William Turner [7]: “ds, ds, represent the decidua serotina of the placenta; t, t trabeculae of serotina passing to the foetal villi; ca curling artery; up utero-placental vein; x a prolongation of maternal tissue on the exterior of the villus outside the cellular layer e', which may represent either the endothelium of the maternal blood vessel, or delicate connective tissue belonging to the serotina, or both. The layer e' represents maternal cells derived from the serotina. The layer of foetal epithelium cannot be seen on the villi of the fully formed human placenta” (original legend). Turner's strands of maternal decidual cells e' moving to the villi are in reality the cytotrophoblastic cell columns of anchoring villi.
3. Huxley's intuition: the importance of the decidua
The term “decidua” was introduced by William Hunter and illustrated in his beautiful ‘Anatomia uteri humani gravidi tabulis illustrata’ (1774) [9]. With this term he described the “membranes” which envelop the conceptus, what we now call the decidua vera and the decidua capsularis, destined to be discarded at parturition (Latin decidere, to fall off). His brother John Hunter also pointed out the existence of a decidua basalis or serotina underneath the placenta. Since he had noticed that in cases of ectopic pregnancies in the human similar “membranes” develop in utero, he concluded that the decidua originates from the uterine mucosa (quoted by De Wit, [10]).
The role of the decidua in placental development remained obscure for a long time. Nevertheless, in the wake of Darwinism, 19th century theorists on evolution saw the presence of this modified maternal tissue layer as a possible clue for understanding the evolution of viviparity. Darwin's close friend Thomas Huxley (1825–1895) considered the presence of a decidua as an indication of the common descent of particular groups of mammals. On this basis he introduced a broad subdivision of mammals into Deciduata and Adeciduata [11]. Elaborating on the topic of decidualization, he reported some important observations on the rat placenta, which reflected an intuitive grasp of its functioning [12]. He correctly identified the maternal arterial channels which run through the whole thickness of the placenta, and pointed out their continuity with a cluster of maternal blood vessels in the uterine wall, or rather the mesometrial triangle which is situated beyond the decidual layer. Still, it is not entirely clear what exactly he was looking for by emphasizing the existence of a decidua in certain groups of mammals. Did he consider this tissue layer as a kind of buffer zone, shielding the uterus from the placenta? The fact that he considered the decidua as the region of placental separation at parturition may hint at a protective or quarantining function, but he never said so explicitly.
Unfortunately, the region which he identified as the rat decidua was in reality the trophospongium, a tissue layer belonging to the fetal placenta. In fact, the decidual layer within the inner myometrium in the rat is so much thinned by the end of pregnancy that it only can be identified by careful microscopic examination of thin tissue sections, a technology still inadequate at the time. Huxley reported his observations in 1864, more than a decennium before Turner [8] published his own interpretations of placental histology. It is interesting that both investigators made the similar mistake of wrongly designating particular trophoblast populations as decidual in, respectively, the rat and the human. Later authors regularly quoted Turner as the originator of the idea that the decidua has a protective function against the eroding action of the placenta. However, in discussing the literature on this question, Haig [13] correctly made the point that there is no evidence that Turner (and we may add Huxley) considered the placenta as an invading tissue. The first clear evidence for invading properties of placental cells emerged from further studies on rodent placentation.
4. Duval's inspiration: the invasion of maternal tissue
Mathias Duval (1844–1907) took up detailed histological studies on placental development in different species in order to acquire a better general understanding of viviparous reproduction. His most influential work probably was “Le placenta des rongeurs” (1891), in which he described in detail the placental development of rabbits, guinea-pigs, rats and especially mice [14]. His mouse work was based on a well-dated series of different pregnancy stages, while his rats were caught “in the wild” (i.e. the slaughterhouses in Paris) and dated approximately by comparison with the mouse specimens. The high impact of Duval's histological studies owed a great deal to his mastery in serial sectioning, a technology that by then had reached a high sophistication. His study resulted in a remarkably dynamic concept of placental development (Fig. 3). At one level, he considered that the progressive outgrowth of allantoic vessels into the placenta leads to a gradual expansion of the maternal–fetal exchange area, resulting in the formation of two tissue layers with a gradually shifting border, one vascularised (the labyrinth) and one without fetal blood vessels (the trophospongium). At a second level, he observed placental tissue penetrating into the decidua, via maternal blood vessels or “sinusoids”, an invasion process by which the placenta “vegetates” into the decidua to capture maternal nutrients. Although his studies were focussed on mice, Duval got this inspiration mainly by studying his relatively limited collection of rat specimens. This species shows a remarkably deep endovascular invasion into spiral arteries, in contrast to the mouse. Nevertheless, Duval thought that he could integrate his rat observations into his concept of placental development in the mouse, by assuming that invasion occurred via maternal “blood sinuses” over the whole width of the placenta. This over-generalization resulted in the erroneous interpretation of the glycogen cells of rodent placentas as captured groups of decidual cells [15].
Fig. 3
Diagram summarizing Duval's interpretation of placental development in mice. (1) Progressive outgrowth of allantoic vessels (All) into the placenta results in formation and gradual expansion of the labyrinth (Lab – yellow arrows) shifting the labyrinthine–trophospongial border. By this process increasingly more nutrients are extracted for the developing fetus. (2) Outgrowth of the trophospongium (Ts) into the decidua results in incorporation of decidua-derived glycogen cell islands (Gl) into the placenta. By this process the placenta “vegetates” into the decidua to capture more maternal nutrients.
Duval's exaggerated extrapolations notwithstanding, he probably was the first to correctly describe trophoblast invasion into maternal arteries. However, he was not the first to have seen such unusual intravascular cells within the lumen of spiral arteries. Twenty years before, in 1870, Carl Friedländer [16] had reported the presence of such cells in “uterine sinuses” of an 8 months pregnant human uterus, mentioning fleetingly that arteries were only rarely observed (!) (Fig. 4). He was unable to decide whether these intravascular cells came from the placenta or from the surrounding maternal tissue, but he reported their occurrence as deep as the inner myometrium. Oddly enough, he considered these cells to be multinuclear. Friedländer reasoned that the intravascular cells with accompanying “fibrin” deposits and newly formed patches of connective tissue – which he interpreted as organized thrombi (thickened intimae!) – must considerably slow down and even interrupt the maternal blood supply to the placenta. He obviously considered flow interruption as an important physiological function of these peculiar cells, and he reasoned that failed vascular plugging might be the cause of intrauterine bleeding and maternal death. The leading German pathologists at that time favoured the idea that the endovascular cells were sloughed off from the maternal vascular wall by the bloodstream, as reviewed several decennia later by Otto Grosser [6], who himself definitely considered them to be trophoblastic.
Fig. 4
Friedländer's [16] illustrations of “uterine sinuses”. “Fig 6. Uterine sinus of the placental site, 8 month's pregnancy. a: musculature; b: thickened homogeneous membrane containing dispersed cells; c: multinuclear granulated cells filling the sinus. Fig 7. Similar structure with beginning organization of thrombus. a: musculature; b: connective tissue surrounding the sinus; c: thickened membrane; d: blood corpuscles and multinuclear cells embedded in a fibrous network; e: young connective tissue, an organized thrombus. Fig 8. Multinuclear cells isolated from such a sinus; in between a few blood corpuscles. Fig 9. Uterine sinus of the placental site 3 months' post partum. a: musculature; b: folded membrane; c: organized thrombus” (translated from the original legend). This figure clearly shows different stages of spiral artery remodelling by invading trophoblast. The “thickened membrane” of his Fig. 7c is the fibrinoid layer with embedded trophoblast, which is partially covered by a thickened intima (e). Endovascular cells (his Fig. 8) are no longer considered as multinuclear.
5. Hubrecht's assertion: the naming of the trophoblast
In his studies of mammalian development, the Dutch embryologist Ambrosius Hubrecht (1853–1915) concentrated on the implantation process, this being the most essential step in the evolution towards viviparity. Highly influenced by Darwin's theories of evolution, he considered comparative embryology as important for elucidating relationships and descent of different species. His work on implantation and early placental development in the hedgehog, published in 1889, has become a classic [17]. He had selected this species under the influence of Huxley's ideas regarding the positioning of insectivores at the base of the mammalian evolutionary tree. In specimens collected very early after implantation he noticed the relationship between early proliferating “epiblast of the blastocyst” and maternal blood lacunae “without an endothelial lining”. A direct role of these future placental cells in the uptake of nutrients from the mother's blood seemed obvious, hence the new name he proposed for these cells: “The first new name of which I want definitely to establish the significance… is the name trophoblast. I propose to confer this name to the epiblast of the blastocyst as far as it has a direct nutritive significance, as indicated by proliferating processes, by immediate contact with maternal tissue, maternal blood, or secreted material”.
Attempts to refine this terminology were made by other investigators. For example in 1903 the American anatomist Charles Sedgwick Minot suggested using the term “trophoderm” for the mature form of these cells in the definitive placenta, but Hubrecht reacted to this proposal with scepticism and the term never caught on. A possible reason for this may be that embryologists never could consider the placenta or its cellular constituents as reaching the status of a mature organ. Therefore the embryonic suffix remained stuck to the mature syncytiotrophoblast of the human placenta (instead of syncytiotrophoderm or syncytiotrophothelium?) and even to the invading extravillous trophoblasts (trophocytes?) which are no longer proliferative.
Being basically an embryologist, Hubrecht focussed mainly on the identification and development of the different extra-embryonic membranes, and did not make further attempts to study in detail the trophoblast–maternal interactions in later stages of placental development. Other investigators increasingly focussed on early post-implantation stages, also in the very rare specimens of human pregnancy obtained after accidental deaths, and confirmed the presence of a trophoblast layer at the interface between maternal and fetal circulations. At the turn of the century the villous coverings of the human placenta were no longer taken to be derived from maternal tissues, although it is difficult to attach the name of a particular investigator to this shift in histological understanding [6].
6. Nitabuch's allegation: the fibrinoid as a barrier to trophoblast invasion
As the last part of the early history of placentology, we consider the alleged association of trophoblast invasion with “Nitabuch's layer”, a term which is still used in the obstetrics literature. The concept is based on some observations recorded in Raissa Nitabuch's doctoral thesis at the University of Bern (1887), written under the promotership of Theodor Langhans, to our knowledge her only publication on the subject [18]. Her material consisted of just one 6 months' pregnant uterus, with the placenta still attached. In her work she mainly concentrated on the connections between the maternal vasculature and the intervillous space. In her introductory chapter, however, she listed a few specific observations on the decidua (“Serotina”), including the presence of a dark line (“ein dunkler Streifen”) that ran close to the basal plate almost over the whole extent of the placenta. This line formed a rather sharp demarcation from the deeper decidua, but was less pronounced towards the superficial layers facing the placenta. She noted that its composition very much resembled the “Fibrin von Langhans”, which her promoter had described in the chorionic plate. The substance was extracellular in nature, but contained “canals”, which regularly enclosed nuclear remnants or “bigger cells”, resembling the dark cells at the placental side rather than the lighter cells at the decidual side of this fibrinoid layer. As to the significance of this layer, Nitabuch quoted Langhans' suggestion that it represented the line of adherence between the “Zellschicht des Chorions” (cytotrophoblastic shell) and the decidua, suggesting a neat border separating the zone containing “chorionic cells” from the deeper decidua. In her thesis she never showed an inclination to attach her name to the “Streifen”, but obviously her Professor must have referred to “Nitabuch's layer” afterwards, as an addition to the fibrinoid he had described himself, and this name has permanently stuck.
While describing maternal arterial cross-sections close to the intervillous space, Nitabuch also noted that breaches occurred in the endothelial layer, which were seemingly replaced by masses of cells, morphologically similar to those she had described at the inside of the fibrinoid layer. Although she did not quote Friedländer's 1870 publication, she almost certainly had hit upon similar endovascular cells, which she seems to have considered as “chorionic”. Unfortunately Nitabuch did not provide detailed histological illustrations of these arteries, or even of “her” fibrinoid layer.
Although the term “Nitabuch's layer” is still generally used, its physiological meaning has always remained uncertain. In their famous textbook, Benirschke and Kaufmann rightly emphasized the variability of this layer in different specimens [19]. The possible importance of Nitabuch's layer was advertised in a speculative paper by Bardawil and Toy (1959) – written in the early days of reproductive immunology – in which they considered it to be an effective barrier preventing maternal immune attack on the semi-allogeneic trophoblastic cells [20]. This idea was obviously based on Nitabuch's description of two different cell populations on the two sides of this layer, which seemed to separate effectively the trophoblast from the decidua. This was long before investigators fully realized the actual depth of trophoblast invasion into the uterine wall.
7. Conclusion
Investigating the historical roots of any field of research, including placentology, reveals science in action. The effort is worthwhile since it deepens our insight into the way science works and why investigators in the past were sometimes led into wrong conclusions. This is particularly exemplified by the shifting concepts about the nature of the fetal–maternal interface in the placenta. Current dogmatic opinions, limited technological possibilities and the lack of appropriate specimens are all factors that may have been responsible for faulty interpretations in the past. Paradigm shifts are periodically necessary for scientific progress, but should not lead to new dogmatism, looking down upon all previous attainments as being obsolete.
A major step forward in 19th century microscopy was the new technology of serially sectioning paraffin-embedded tissues, allowing 3D reconstruction of complex histological structures. The next important breakthrough for understanding morphology was the development of sophisticated cell identification techniques during the last decennia of the previous century. In applying such modern techniques, however, we should not ignore the painstaking work executed by our predecessors, on whose shoulders we now stand. Also new histological data based on distributions of marker molecules need to fit within histological continuity and shifting cell distributions in space and time. An acquaintance with the findings as well as the previous erroneous interpretations of “classical” morphology remains essential for keeping a critical mind on our present understanding of the placenta. In this perspective we quote the late S.J. Gould: “I do not trace the history of this canonical line in the interests of antiquarian pedantry, but because the enterprise can yield such rich rewards in teaching us about crucially important, and often unrecognized, biases in our modes of thought and styles of storytelling” [21]. Maybe the most important lesson learnt from the past is that even in our current thinking we might be wrong, or at least not completely right.
8. Conflict of interest
The authors do not have any potential or actual personal, political, or financial interest in the material, information, or techniques described in this paper.
Acknowledgement
The authors wish to thank Tom Pijnenborg for restoring the faded ancient prints of Fig. 1, Fig. 2, Fig. 4.
References
- . On the structure of the placenta. Reprint of the original 1786 edition, with notes by Richard Owen Observations on certain parts of the animal economy. Philadelphia: Haswell, Barrington & Haswell; 1840;93–103
- . Chapter 38: of the oxygenation of the blood in the lungs, and in the placenta. Chapter 39: of generation In: Zoonomia; or the laws of organic life. 3rd ed.. London: J. Johnson; 1801;
- . Erasmus Darwin's enlightened views on placental function. Placenta. 2007;28:775–778
- . Exploring the placental maze: the development of our knowledge of the relation between the bloodstreams of mother and infant in utero. American Journal of Obstetrics and Gynecology. 1963;86:408–418
- . Principles of comparative physiology. 4th ed.. London: Churchill; 1854;p. 626–7
- . Frühentwicklung, Eihautbildung und Placentation des Menschen und der Säugetiere. München: J.F. Bergmann; 1927;
- . Sir William Turner and his studies on the mammalian placenta. Australian and New Zealand Journal of Surgery. 2003;73:449–452
- . Some general observations on the placenta, with especial reference to the theory of evolution. The Journal of Anatomy and Physiology. 1877;11:33–53
- . Anatomia uteri humani gravidi tabulis illustrata. The anatomy of the human gravid uterus exhibited in figures, Birmingham, Alabama: Gryphon Editions; 1980;Facsimile edition of the original edition (Birmingham: John Baskerville 1774)
- . An historical study on theories of the placenta to 1900. J History of Medicine and Allied Sciences. 1959;14:360–374
- . On the classification of animals. The subdivisions of the mammalia larger than orders. Lectures on the elements of comparative anatomy. London: Churchill; 1864;p. 87–112
- . Thomas Huxley and the rat placenta in the early debates on evolution. Placenta. 2004;25:233–237
- . William Turner's lectures on the comparative anatomy of the placenta. Placenta. 1997;18:371–374
- . Le placenta des rongeurs. Paris: Felix Alcan; 1892;p. 640
- . Mathias Duval on placental development in mice and rats. Placenta. 2006;27:109–118
- . Physiologisch-anatomische Untersuchungen über den Uterus. Leipzig: Simmel; 1870;p. 31–6
- . Studies in mammalian embryology. I. The placentation of Erinaceus europaeus, with remarks on the phylogeny of the placenta. Q J Microsc Sci. 1889;30:283–404
- . Beiträge zur Kenntniss der menschlichen Placenta. Bern: Stämpfli'sche Buchdruckerei; 1887;
- . Pathology of the human placenta. 3rd ed.. New York, Berlin, London: Springer-Verlag; 1995;p. 223–6
- . The natural history of choriocarcinoma: problems of immunity and spontaneous regression. Annals of the New York Academy of Sciences. 1959;80:197–261
- . A special fondness for beetles. Dinosaur in a haystack. London: Jonathan Cape; 1996;p. 377–87
PII: S0143-4004(07)00232-9
doi:10.1016/j.placenta.2007.09.006
© 2007 IFPA and Elsevier Ltd. All rights reserved.
