JPM Vol. 28, No. 2 September 1992:61-66
REVIEWS
Placental Perfusion An Overview of the Literature Daniela Perinatal Clinical Pharmacology
Omarini,
Vanna
Pistotti,
and Maurizio
Bonati
Unit (D.O., M.B.), and Gustavus A. Pfeiffer Memorial Library, Istituto di Recerche Farmacologiche “Mario Negri,” Milano, Italy
Placental perfusion techniques are currently used to study not only the organ functions but also the transfer profile and metabolic pathway of different compounds. In view of the interest in the mechanism of transfer and potential adverse effects of compounds there are numerous publications on the topic, but no systematic picture is yet available. Thus an overview has been made of all studies published from 1966 to 1990 that use this experimental approach. Out of 359 computer-retrieved articles, 266 (74%) actually dealt with the target topic; 68 articles were added after a systematic hand search, so a total of 334 articles were analyzed. The distribution of papers per year was constant until 1980, and rose significantly thereafter. Animal studies using placental perfusion were performed either in situ or in vitro, whereas human investigations were mostly examined by in vitro perfusion techniques. Animal experiments were done on seven species, the guinea pig being the most widely used. The aims of all studies could be divided into five main categories: 132 studies researched the kinetics of compounds in the placenta; 100 studies investigated placental metabolism; methodology of perfusion was reported in 22 articles; and 49 studies examined physiological changes of placental variables. A clear increase in pharmacological studies was noted starting from 1986 (there were 31 such studies). Compounds studied were either endogenous or exogenous. Almost all endogenous compounds were investigated, some of them quite extensively (mainly hormones, angiotensin, glucose, and lactate). There seemed to be no preferential field for exogenous compounds (62 compounds could be assigned to 20 classes). A total of 95 journals were represented, but only two of them (American Journal of Obstetrics and Gynecology and Placenta) presented any continuity in the field. The 334 papers analyzed were published by 591 authors, but 78% of all articles were written by 107 scientists, from six main groups. It appears that despite the wide-ranging approach and the large amount of work done using these experimental techniques, the real role of placental perfusion as a biomedical research tool has yet to be defined. Keywords: ledon.
Placental perfusion techniques;
Introduction The placenta is the crucial link between the mother and her fetus, and its functional capacity has far-reaching consequences not only for intrauterine life, but also for the whole of postnatal development.
Address reprint requests to Dr. D. Omarini, Perinatal Clinical Pharmacology Unit, Istituto di Ricerche Farmacologiche ‘Mario Negri,’ Via Eritrea 62, 20157 Milano, Italy. Received March 1992; revised and accepted June 1992. Journal of Pharmacological and Toxicological Methods 28, 61-66 (1992) 0 1992 Elsevier Science Publishing Co., Inc., 655 Avenue of the Americas, New York, NY 10010
Placental metabolism;
Human placental
coty-
Because of the difficulty of access to the placenta, human research has been limited to morphological or biochemical investigations on organs obtained from abortions or at delivery. Only in the last few decades have methods and techniques been developed to obtain broader information on the functional performance of the placenta throughout pregnancy (Miller, 1986). We now know that the early theory of the placental barrier is only of limited and anecdotical value, as many studies clearly showed that a wide class of exoge-
62
JPM Vol. 28, No. 2 September 1992:61-66
nous and endogenous compounds are transferred and can affect the embryo or fetus (Mihaly and Morgan, 1984; Schneider, 1991). However, information about the functional processes involved is still scanty. Besides technical limitations, concern about the safety of the mother and her fetus and legal restrictions have rightly to be added. Thus, in view of the widely recognized importance of this field of interest, research has had to be limited to different animal species. Nowadays one of the main techniques for investigating placental transfer and function is placental perfusion, particularly in situ and in vitro (ex vivo) in animals, and in vitro on human specimens. The in vitro approach is a method for artificial perfusion of the fetal capillaries and, in most cases, of the intervillous space simultaneously, using an isolated placenta or frequently only a single cotyledon (Brandes et al., 1983; Carstensen et al., 1977; Contractor and Stannard, 1983; Penigel, 1968; Penfold et al., 198 1; Schneider and Huch, 1985). In situ approach involves perfusion of the fetal compartment of an intact placenta by cannulation of umbilical vessels (Shaw et al., 1990; Stulc and Stulcova, 1986; Hamshaw-Thomas and Reynolds, 1985; Reynolds and Young, 1971). Many investigations have employed placental perfusion, and the interest in the mechanism of transfer and potential adverse effects of compounds has generated much information. However, the fact that no systematic picture is available offered a challenging opportunity for an overview of studies published from 1966 to 1990.
Materials and Methods Rather than just review existing ones, we decided to seek out the primary sources of information as far as possible, in the hope that this effort would not only yield more reliable data, but also provide an opportunity for reconsidering methodological problems. An English-language search was conducted by MEDLINE from 1966 through 1990, and a manual search was made, with a check of references quoted in main original articles and reviews, on published studies using placental perfusion as the experimental method.’ Books and proceedings from meetings and congresses were not considered. The strategy adopted for the online search and the results is summarized in Table 1. A total of 359 articles were computer retrieved, and 266 (74%) dealt with the target topic, after exclusion of nonpertinent articles (i.e., perfusion of liver, kidney, lung, etc.) or letters, short communications, and a
r The detailed list of all 334 analyzed papers will be mailed if requested by members.
Table 1. Procedure
for the
MEDLINE
Online Search
1966-1990 Query No.
Search Terms
#I
Maternal-Fetal Exchange” or Biological Transporta or In-Vitro” Placenta$b samePerfus$b Placenta0 and Perfusion” #2 or #3 #I and #4 #S and English language
#2 #3 #4 #5 #6
No. of References 246. 537 740 353 741 439 359
a MeSH (Medical Subject Headings) Descriptors. b Text word in the abstract and title fields only.
monograph. Another 68 papers were identified by hand-search suggesting that abstracting indexers of MEDLINE in this area are not unduly reliable, in comparison with the hand-search. Thus 334 papers, written in English language, published over the 1966-1990 period, constituted the material for the analysis. Data were processed by SAS (Statistical Analysis Software) programs on an IBM 43-41 mainframe. The aim of the study was, first, to highlight any trends in the literature over the years, considering the main information: year of publication, authors, and journals. Second, a content analysis of all 334 papers was attempted, in which five main variables were examined (animal species, type of model, in vitro or in situ, study aims, and compounds investigated), so that one could draw a general picture of the state of the art.
Results The 334 articles were published by 591 authors. Of these, 400 appeared in only one paper each (covering 67% of the total presences); 84 authors appeared on two papers each (14% of all authors). Only 107 authors (18%) appeared in at least three papers each, but 78% of the literature in this field was written by these scientists. Nearly all of them can be traced to a few groups constantly present in this literature field. They are from New York (U.S.A.), Bern (Switzerland), Paris (France), London (England), Prague (formerly Czechoslovakia), Hamburg (Federal Republic of Germany), and Melbourne (Australia). These groups are made up mainly of the 42 authors (7% of all present) of the 55% of all the published studies using placental perfusion as the main investigation method. The 334 papers appeared in a total of 95 journals. Of these, 45 published only one paper each in the period considered, 17 published two papers, 12 published three papers, and four published four papers each. Seventeen journals published at least five papers each (for
63
OMARINI ET AL. PLACENTAL PERFUSION
of them concerned perfusion (in situ or in vitro) of intact human fetoplacental units, and 20 performed in situ perfusion on pregnant women. Twenty-eight studies (8%) used in vitro perfusion of animal placenta, and six were sporadic presentations of research utilizing animal and human placenta or in vitro and in situ techniques in the same study. These last 34 papers are evenly distributed over the period considered, without any particular pattern. Figure 2 shows the trend of placental perfusion experiments over years. This first analysis showed that the distribution of human placental studies was fully comparable to the frequency distribution of in vitro experiments and the same can be seen for animal and in situ experiments. These two subgroups constitute 80% of the literature. The trend over the years was confirmed by statistical analysis, because when applying a polynomial regression of second order to all data, a clear differ-
a total of 203 papers), but only three were mainly involved: American Journal Obstetrics and Gynecology (32 papers), Placenta (31 papers), and Acta Endocrinologica (22 papers). The majority of studies used the human placenta (221 papers, 66%). There were 109 (33%) studies on animal placenta, whereas both human and animal species were the topics of four papers (Figure 1 (a,b)]. The most frequently used method was in vitro perfusion (213 papers, 64%). In human studies, placentas were obtained either after vaginal delivery or after cesarean section, whereas surgically isolated whole placentas were commonly used in animal experiments. The in situ approach was reported in 103 papers (3 1%). Both in vitro and in situ techniques were utilized in six studies (2%). Out of the 334 studies, 32 papers reported methodology that has been completely forgotten since 1973: 12
20 m
in situ
24
a
on fetchplacental
22 3
EI in vitro
26
unit
20 10 16 14 12 10
a 6 4 2 0
Figure 1. (a) Distribution of human placenta perfusion techniques over years. (b) Distribution of animal placenta perfusion techniques over years.
66 67 66 69 70 71 72 73 74 75 76 77 70 79 00 01 02 83 04 85 86 67 00 09 90
year
A 20 26 24 -
n q
in situ in vitro
22 20 1816 14 1210 0 6 4 2 0 66 67 60 69 70 71 72 73 74 75 76 77 70 79 80 01 82 03 84 85 86 07 08 89 90
year
64
JPM Vol.28,No. 2 September 1992:61-66
2
30-
0
a, ma Q
25.
5 2
20.
El 3 g .k
15.
10.
5.
0 OJ/
,,,,,,,,,,,,,,,,,,,,,,,, 66
68
70
72
74
76
70
80
82
84
86
88
90
year Figure 2. Growth of literature
on placental perfusion during
investigated endogenous compounds were hormones, particularly sex hormones (androgens and estrogens), prostaglandins, and steroids. Angiotensin was investigated in 12 studies (11 of them on human placenta in vitro). Glucose and lactate were the most widely investigated carbohydrates, with oxygen (eight studies) and calcium (14 studies), the most studied elements. All other endogenous substances were widely distributed among enzymes, proteins, vitamins and coenzymes, aminoacids, lipids (particularly free fatty acids), and nucleic acids, without any marked favorites. Excluding antipyrine and creatinine from the analysis because of their common use as markers for compound transfer in placental perfusion experiments, the remaining 62 exogenous compounds could be assigned to 20 classes. Only phenytoin, bupivacaine, pethidine, ritodrine, diazepam, and ethanol were studied three times each.
the period from 1966 to 1990.
Discussion ence can be seen during the period between 1966 and 1973 (when two techniques disappeared: the one in situ on pregnant women and the other on the fetoplacental unit, with a nearly constant production of papers) and 1974 and 1990 (showing the effect of the appearance of the new method for perfusion of a human placental cotyledon in vitro and the rapid subsequent growth in frequency of papers). The use of animal models has remained more or less constant over the years. Animal experiments have been done on seven species. The guinea pig was the animal most used (71 studies), followed by the sheep (13 studies), rabbit and rat (10 each), the monkey (Macaca Mulatta, two), goat and pig (one each). One study employed both the sheep and rabbit. Five main topics of study were identified, with a different frequency distribution over the period reviewed. The transfer of substances across the placental barrier was investigated in 132 papers (39%). Metabolism of compounds in the placenta was reported in 100 papers (30%). The third group comprised 22 papers (7%) that discussed the placenta perfusion technique. In 3 1 papers (9%), the effects of drugs on the placenta were studied, and 49 studies (15%) examined physiological changes in placental variables such as blood flow. The frequency distribution of the different topics over the whole period was examined to identify any trend in the type of studies. In the last 5 years we could identify a change in the distribution: only live out of 110 studies reported the methodology of perfusion, and a clear increase in pharmacological studies can be observed starting in 1986. Compounds studied by this method were 196, that is, 132 endogenous and 64 exogenous. The most widely
Analysis of the 334 papers related to placental perfusion techniques offers an interesting insight into the main aspects of this research method. A minority of authors account for about 80% of the reported literature, and nearly all these scientists come from a few groups, whose senior scientists are among those who have worked longest in the field. This can be explained probably by the fact that working with placental perfusion methods calls for a great expenditure of technical details and time. A lot of physiological, biochemical, physical, and mathematical elements are involved and, therefore, for a structured group to follow this field over the years requires much effort, and not the least, tinancial. A parallel situation can be noted for journals. Placenta, a periodical published for the first time in 1980, and the American Journal of Obstetrics and Gynecology, which has the longest tradition and continuity in this field, both present contents closely related to the field of research. A peculiar position is occupied by Acta Endocrinologica, which published 22 papers, reflecting the fact that the most frequently examined substances in these reports were hormones. The techniques mainly used in these studies were perfusion on pregnant women and the human fetoplacental unit. Until 1973, 17 of the 22 papers were published by the same author. Both the author and the techniques subsequently vanished and this journal published only three more papers concerning studies using placental perfusion. A recurrent question was how closely the in vitro perfused placenta resembles the intrauterine placenta and how the performance of the in vitro preparations could be affected by traumatization of the structure and metabolic integrity following delivery.
OMARINIETAL. PLACENTAL
PERFUSION
There is ample evidence that despite all abuse the perfused placenta .retains many complex functions (e.g., active and energy-dependent transport of amino acids across the placental barrier by a stereospecific transport system that establishes a gradient towards the fetus; facilitated transport of o-glucose, which is three times higher than the diffusion rate of the L-isomer; and production of several placental peptides) (Dancis, 1985). A main consideration in this context is that the mother and fetus continuously modify placental “perfusates” in vivo. By removing the mother and fetus from the system, the perfusionist achieves a desirable simplicity and control of experimental conditions, so the confounding effects of maternal and fetal metabolism are avoided, and placental transfer and metabolism can be studied in isolation, Another recurrent question was whether the human or animal placenta is to be preferred, but there appears to be no clear answer. For practical, ethical, and legal reasons nowadays this question is closely related to the choice of using an in vitro or in vivo approach. There are some advantages to the in vivo approach, but such experiments are limited by concerns for the safety of mother and infant. These advantages are physiological, for example, original blood instead of highly artificial perfusates, no structural damage following separation from the uterine wall, and no anoxic period. Probably these were reasons why most researchers doing animal experiments prefer in vivo approaches and this also explains why the in situ animal perfusion technique has been more or less constant throughout the years. The fact that the guinea pig was the most widely selected species surely reflects the similarity of the hemochorial placenta to the human one, and the same reason is valid for the rabbit. An exceptional position is held by the sheep because most experimental work on placental hemodynamics has been done on this animal. Although there are fewer parallels between the epitheliochorial placenta and the human hemochorial one, in the sheep it is possible to perform intricate surgery and take fetal blood samples large enough for chemical analysis. It emerges clearly, however, that there is no ideal experimental animal for intrauterine surgery whose placenta is functionally and completely similar to the human placenta and fetuses are large enough to permit repeated sampling. It must also be borne in mind that the placenta is the mammalian organ that shows the greatest variations in structure among species (Faber and Thornburg, 1983). Differences in function must, therefore, also be assumed, and several important ones have already been documented. This statement is obviously reflected in the finding that in vitro perfusion of human placental
65
tissues has risen in the last year, whereas animal observations are more or less constant. Although attempts have been made to perfuse the whole placenta, methodological problems are such that nowadays this approach is nearly forgotten. Today many researchers follow the original protocol developed by Panigel (1968) or a modified version, perfusing a single cotyledon of a human placenta, with the maternal surface uppermost. An English group has developed a technique to perfuse the placenta with the maternal surface downwards, which seems to shorten the time between delivery and reperfusion (a crucial point in this technique) and avoids pooling maternal side effluents. Because these two approaches have never been compared in the same laboratory as far as we know, no distinct advantages of either one or the other version have been documented. With regard to the different topics that have been investigated with placental perfusion, most experiments were concerned with the transfer of substances across the placental barrier. Most of these investigated carbohydrates (particularly glucose as one of the most important fuels for the fetus), almost all known amino acids, and fatty acids. Many set out not only to answer the question whether these substances cross the placental barrier, but also to assess quantitative parameters. Metabolic reactions of the different substances constitute another main topic, and as a result of these experiments it has been demonstrated, for example, that the placenta is a metabolically active organ that has its own requirements because it consumes appreciable amounts of glucose and oxygen. Other experiments investigated the transfer of drugs, and the results are particularly valuable nowadays as almost all pregnant women receive some drug therapy. The fact that pharmacological perfusion experiments have increased in the last years reflects the increased sensitivity towards problems related to drug use in pregnancy. Some methodological problems emerged in reviewing all these articles. Despite the impressive amount of data, it remains very difficult for anyone to compare the results of studies with the same declared aim (i.e., studies on the same compounds, on similar physiological variables), because many papers fail to report some basic information. Kelmann and van Kreel (1987) proposed a few general criteria useful for judging the viability of placental perfusions in animals. Extending these criteria to perfusions of human placentas would provide a valuable basis for comparisons. In most of the papers (with the exception of declared “methodological” articles), information on the perfusion equipment were scant, and often even the indications for “open” or “closed” circulation were missing.
66
JPM Vol. 28, No. 2 September 1992:61-66
We again suggest that efforts should be made to report the main information to permit a comparisons of results. Examples are 1) measurements of maternal and fetal flows with related pressures, which could help one understand and minimize artifacts due to changes in permeability during perfusion; 2) possible maternal flow to the placenta in in vivo experiments; 3) if a drug or chemical agent was used during perfusion (anesthetics, heparin, antibiotics, antifoam, etc.); and 4) the composition of the perfusion medium and the pH measurements of flow in and out of the placenta (especially for pharmacological studies). It would also be useful if the variables-listed as a “perfusion index” (e.g., antipyrine, creatinine, and inulin transfer, oxygen balante)-were reported using the same international unit in different studies, as sometimes it is impossible to convert one unit to another. If the common intent is to provide a valid method to study more complex questions (for instance, effects of pharmacological substances on placental behavior) and, in general, study how compounds reach the fetus and their effects on the physiology of the placenta (and, therefore, on pregnancy), then the points briefly mentioned above constitute the minimum basis to enable other people working in the same field to use these studies as a constructive source of information. Even though an impressive amount of work has been done with these techniques, the approach is still more piecemeal than systematic, and, therefore, the full potential of this approach in biomedical research is far from being fully exploited.
of glucose, 800-806.
inulin and antipyrine.
Am .I Obstet Gynecol
146:
Carstensen M, Leichtweiss HP, Molsen G, Schroder H (1977) Evidence for a specific transport of o-hexoses across the human term placenta in vitro. Arch Gynaek 2221187-196. Contractor SF, Stannard PJ (1983) The use of AIB transport to assess the suitability of a system of human placental perfusion for drug transport studies. Placenta 4: 19-30. Dancis J (1985) Why perfuse the human placenta. In In Vitro Perfusion of Human Placental Tissue (vol 13 of Contribution to Gynecology and Obstetrics). Eds., H. Schneider and J. Dancis. Basel: Karger pp. l-4. Faber JJ, Thomburg KL (1983) Structural features of placental exchange. In Placental Physiology: Structure and Function ofFetomaternal Exchange. Eds., JJ Faber and KL Thornburg. New York: Raven Press, Inc, pp. l-32. Hamshaw-Thomas A, Reynolds F (1985) Placental transfer of bupivacaine, pethidine and lignocaine in the rabbit. Effect of umbilical flow rate and protein content. Br J Obstet Gynecol 92:706713.
Kelman BJ, van Kreel BK (1987) Criteria for standardizing and judging viability placental perfusions in animals. Trophoblast Research 2:515-522.
Mihaly GW, Morgan DJ (1984) Placental drug transfer: Effects of gestational age and species. Pharamcol Ther 23:253-266. Miller RK (1986) Techniques for the study of placental function. In Drug and Chemical Action in Pregnancy: Pharmacologic and Toxicologic Principles. Eds., S Fabro and AR Scialli. New York:
Marcel Dekker, Inc, pp. 128-135. Panigel M (1968) Placental perfusion. In Fetal Homeostasis, New York: Appleton Century Crofts, pp. 15-25.
vol 4.
Penfold P, Drury L, Simmonds R, Hytten FE (1981) Studies of a single placental cotyledon in vitro. I. The preparation and its viability. Placenta 2:149-154. Reynolds ML, Young M (1971) The transfer of free a-amino nitrogen across the placental membrane in the guinea-pig. J Physiol214: (Land) 583-597.
The authors thank Dr. Wolfgang Reiber for a critical reading of the manuscript, Lucia August0 Casaroli for computer assistance, Judy Baggott for the editing, and Daniela Miglio for the secretarial help. The generous contribution of the Banca Popolare, Milan, Italy, is also gratefully acknowledged.
Schneider H (1991) Placental transport function. Reprod Fertil Dev 3:345-353.
Schneider H, Huch A (1985) Dual in vitro perfusion of an isolated lobe of human placenta: Method and instrumentation. In In vitro Perfusion of Human Placental Tissue (vol 13 of Contribution to Gynecology and Obstetrics). Eds., H Schneider and J Dancis. Basel: Karger pp. 40-45.
References
Shaw AJ, Mughal MZ, Mohammed T, Maresh MJA, Sibley CP (1990) Evidence for active matemofetal transfer of magnesium across the in situ perfused rat placenta. Pediatr Res17(6):
Brandes JM, Tavolini N, Potter BJ, Sarkozi L, Shepard MD, Berk PD (1983) A new recycling technique for human placental cotyledon perfusion: Application to studies of the fetomaternal transfer
Stulc J, Stulcova B (1986) Transport of calcium by the placenta of the rat. J Physiol (Lond) 371:1-16.
622-625.
REVIEWS
Placental Perfusion An Overview of the Literature Daniela Perinatal Clinical Pharmacology
Omarini,
Vanna
Pistotti,
and Maurizio
Bonati
Unit (D.O., M.B.), and Gustavus A. Pfeiffer Memorial Library, Istituto di Recerche Farmacologiche “Mario Negri,” Milano, Italy
Placental perfusion techniques are currently used to study not only the organ functions but also the transfer profile and metabolic pathway of different compounds. In view of the interest in the mechanism of transfer and potential adverse effects of compounds there are numerous publications on the topic, but no systematic picture is yet available. Thus an overview has been made of all studies published from 1966 to 1990 that use this experimental approach. Out of 359 computer-retrieved articles, 266 (74%) actually dealt with the target topic; 68 articles were added after a systematic hand search, so a total of 334 articles were analyzed. The distribution of papers per year was constant until 1980, and rose significantly thereafter. Animal studies using placental perfusion were performed either in situ or in vitro, whereas human investigations were mostly examined by in vitro perfusion techniques. Animal experiments were done on seven species, the guinea pig being the most widely used. The aims of all studies could be divided into five main categories: 132 studies researched the kinetics of compounds in the placenta; 100 studies investigated placental metabolism; methodology of perfusion was reported in 22 articles; and 49 studies examined physiological changes of placental variables. A clear increase in pharmacological studies was noted starting from 1986 (there were 31 such studies). Compounds studied were either endogenous or exogenous. Almost all endogenous compounds were investigated, some of them quite extensively (mainly hormones, angiotensin, glucose, and lactate). There seemed to be no preferential field for exogenous compounds (62 compounds could be assigned to 20 classes). A total of 95 journals were represented, but only two of them (American Journal of Obstetrics and Gynecology and Placenta) presented any continuity in the field. The 334 papers analyzed were published by 591 authors, but 78% of all articles were written by 107 scientists, from six main groups. It appears that despite the wide-ranging approach and the large amount of work done using these experimental techniques, the real role of placental perfusion as a biomedical research tool has yet to be defined. Keywords: ledon.
Placental perfusion techniques;
Introduction The placenta is the crucial link between the mother and her fetus, and its functional capacity has far-reaching consequences not only for intrauterine life, but also for the whole of postnatal development.
Address reprint requests to Dr. D. Omarini, Perinatal Clinical Pharmacology Unit, Istituto di Ricerche Farmacologiche ‘Mario Negri,’ Via Eritrea 62, 20157 Milano, Italy. Received March 1992; revised and accepted June 1992. Journal of Pharmacological and Toxicological Methods 28, 61-66 (1992) 0 1992 Elsevier Science Publishing Co., Inc., 655 Avenue of the Americas, New York, NY 10010
Placental metabolism;
Human placental
coty-
Because of the difficulty of access to the placenta, human research has been limited to morphological or biochemical investigations on organs obtained from abortions or at delivery. Only in the last few decades have methods and techniques been developed to obtain broader information on the functional performance of the placenta throughout pregnancy (Miller, 1986). We now know that the early theory of the placental barrier is only of limited and anecdotical value, as many studies clearly showed that a wide class of exoge-
62
JPM Vol. 28, No. 2 September 1992:61-66
nous and endogenous compounds are transferred and can affect the embryo or fetus (Mihaly and Morgan, 1984; Schneider, 1991). However, information about the functional processes involved is still scanty. Besides technical limitations, concern about the safety of the mother and her fetus and legal restrictions have rightly to be added. Thus, in view of the widely recognized importance of this field of interest, research has had to be limited to different animal species. Nowadays one of the main techniques for investigating placental transfer and function is placental perfusion, particularly in situ and in vitro (ex vivo) in animals, and in vitro on human specimens. The in vitro approach is a method for artificial perfusion of the fetal capillaries and, in most cases, of the intervillous space simultaneously, using an isolated placenta or frequently only a single cotyledon (Brandes et al., 1983; Carstensen et al., 1977; Contractor and Stannard, 1983; Penigel, 1968; Penfold et al., 198 1; Schneider and Huch, 1985). In situ approach involves perfusion of the fetal compartment of an intact placenta by cannulation of umbilical vessels (Shaw et al., 1990; Stulc and Stulcova, 1986; Hamshaw-Thomas and Reynolds, 1985; Reynolds and Young, 1971). Many investigations have employed placental perfusion, and the interest in the mechanism of transfer and potential adverse effects of compounds has generated much information. However, the fact that no systematic picture is available offered a challenging opportunity for an overview of studies published from 1966 to 1990.
Materials and Methods Rather than just review existing ones, we decided to seek out the primary sources of information as far as possible, in the hope that this effort would not only yield more reliable data, but also provide an opportunity for reconsidering methodological problems. An English-language search was conducted by MEDLINE from 1966 through 1990, and a manual search was made, with a check of references quoted in main original articles and reviews, on published studies using placental perfusion as the experimental method.’ Books and proceedings from meetings and congresses were not considered. The strategy adopted for the online search and the results is summarized in Table 1. A total of 359 articles were computer retrieved, and 266 (74%) dealt with the target topic, after exclusion of nonpertinent articles (i.e., perfusion of liver, kidney, lung, etc.) or letters, short communications, and a
r The detailed list of all 334 analyzed papers will be mailed if requested by members.
Table 1. Procedure
for the
MEDLINE
Online Search
1966-1990 Query No.
Search Terms
#I
Maternal-Fetal Exchange” or Biological Transporta or In-Vitro” Placenta$b samePerfus$b Placenta0 and Perfusion” #2 or #3 #I and #4 #S and English language
#2 #3 #4 #5 #6
No. of References 246. 537 740 353 741 439 359
a MeSH (Medical Subject Headings) Descriptors. b Text word in the abstract and title fields only.
monograph. Another 68 papers were identified by hand-search suggesting that abstracting indexers of MEDLINE in this area are not unduly reliable, in comparison with the hand-search. Thus 334 papers, written in English language, published over the 1966-1990 period, constituted the material for the analysis. Data were processed by SAS (Statistical Analysis Software) programs on an IBM 43-41 mainframe. The aim of the study was, first, to highlight any trends in the literature over the years, considering the main information: year of publication, authors, and journals. Second, a content analysis of all 334 papers was attempted, in which five main variables were examined (animal species, type of model, in vitro or in situ, study aims, and compounds investigated), so that one could draw a general picture of the state of the art.
Results The 334 articles were published by 591 authors. Of these, 400 appeared in only one paper each (covering 67% of the total presences); 84 authors appeared on two papers each (14% of all authors). Only 107 authors (18%) appeared in at least three papers each, but 78% of the literature in this field was written by these scientists. Nearly all of them can be traced to a few groups constantly present in this literature field. They are from New York (U.S.A.), Bern (Switzerland), Paris (France), London (England), Prague (formerly Czechoslovakia), Hamburg (Federal Republic of Germany), and Melbourne (Australia). These groups are made up mainly of the 42 authors (7% of all present) of the 55% of all the published studies using placental perfusion as the main investigation method. The 334 papers appeared in a total of 95 journals. Of these, 45 published only one paper each in the period considered, 17 published two papers, 12 published three papers, and four published four papers each. Seventeen journals published at least five papers each (for
63
OMARINI ET AL. PLACENTAL PERFUSION
of them concerned perfusion (in situ or in vitro) of intact human fetoplacental units, and 20 performed in situ perfusion on pregnant women. Twenty-eight studies (8%) used in vitro perfusion of animal placenta, and six were sporadic presentations of research utilizing animal and human placenta or in vitro and in situ techniques in the same study. These last 34 papers are evenly distributed over the period considered, without any particular pattern. Figure 2 shows the trend of placental perfusion experiments over years. This first analysis showed that the distribution of human placental studies was fully comparable to the frequency distribution of in vitro experiments and the same can be seen for animal and in situ experiments. These two subgroups constitute 80% of the literature. The trend over the years was confirmed by statistical analysis, because when applying a polynomial regression of second order to all data, a clear differ-
a total of 203 papers), but only three were mainly involved: American Journal Obstetrics and Gynecology (32 papers), Placenta (31 papers), and Acta Endocrinologica (22 papers). The majority of studies used the human placenta (221 papers, 66%). There were 109 (33%) studies on animal placenta, whereas both human and animal species were the topics of four papers (Figure 1 (a,b)]. The most frequently used method was in vitro perfusion (213 papers, 64%). In human studies, placentas were obtained either after vaginal delivery or after cesarean section, whereas surgically isolated whole placentas were commonly used in animal experiments. The in situ approach was reported in 103 papers (3 1%). Both in vitro and in situ techniques were utilized in six studies (2%). Out of the 334 studies, 32 papers reported methodology that has been completely forgotten since 1973: 12
20 m
in situ
24
a
on fetchplacental
22 3
EI in vitro
26
unit
20 10 16 14 12 10
a 6 4 2 0
Figure 1. (a) Distribution of human placenta perfusion techniques over years. (b) Distribution of animal placenta perfusion techniques over years.
66 67 66 69 70 71 72 73 74 75 76 77 70 79 00 01 02 83 04 85 86 67 00 09 90
year
A 20 26 24 -
n q
in situ in vitro
22 20 1816 14 1210 0 6 4 2 0 66 67 60 69 70 71 72 73 74 75 76 77 70 79 80 01 82 03 84 85 86 07 08 89 90
year
64
JPM Vol.28,No. 2 September 1992:61-66
2
30-
0
a, ma Q
25.
5 2
20.
El 3 g .k
15.
10.
5.
0 OJ/
,,,,,,,,,,,,,,,,,,,,,,,, 66
68
70
72
74
76
70
80
82
84
86
88
90
year Figure 2. Growth of literature
on placental perfusion during
investigated endogenous compounds were hormones, particularly sex hormones (androgens and estrogens), prostaglandins, and steroids. Angiotensin was investigated in 12 studies (11 of them on human placenta in vitro). Glucose and lactate were the most widely investigated carbohydrates, with oxygen (eight studies) and calcium (14 studies), the most studied elements. All other endogenous substances were widely distributed among enzymes, proteins, vitamins and coenzymes, aminoacids, lipids (particularly free fatty acids), and nucleic acids, without any marked favorites. Excluding antipyrine and creatinine from the analysis because of their common use as markers for compound transfer in placental perfusion experiments, the remaining 62 exogenous compounds could be assigned to 20 classes. Only phenytoin, bupivacaine, pethidine, ritodrine, diazepam, and ethanol were studied three times each.
the period from 1966 to 1990.
Discussion ence can be seen during the period between 1966 and 1973 (when two techniques disappeared: the one in situ on pregnant women and the other on the fetoplacental unit, with a nearly constant production of papers) and 1974 and 1990 (showing the effect of the appearance of the new method for perfusion of a human placental cotyledon in vitro and the rapid subsequent growth in frequency of papers). The use of animal models has remained more or less constant over the years. Animal experiments have been done on seven species. The guinea pig was the animal most used (71 studies), followed by the sheep (13 studies), rabbit and rat (10 each), the monkey (Macaca Mulatta, two), goat and pig (one each). One study employed both the sheep and rabbit. Five main topics of study were identified, with a different frequency distribution over the period reviewed. The transfer of substances across the placental barrier was investigated in 132 papers (39%). Metabolism of compounds in the placenta was reported in 100 papers (30%). The third group comprised 22 papers (7%) that discussed the placenta perfusion technique. In 3 1 papers (9%), the effects of drugs on the placenta were studied, and 49 studies (15%) examined physiological changes in placental variables such as blood flow. The frequency distribution of the different topics over the whole period was examined to identify any trend in the type of studies. In the last 5 years we could identify a change in the distribution: only live out of 110 studies reported the methodology of perfusion, and a clear increase in pharmacological studies can be observed starting in 1986. Compounds studied by this method were 196, that is, 132 endogenous and 64 exogenous. The most widely
Analysis of the 334 papers related to placental perfusion techniques offers an interesting insight into the main aspects of this research method. A minority of authors account for about 80% of the reported literature, and nearly all these scientists come from a few groups, whose senior scientists are among those who have worked longest in the field. This can be explained probably by the fact that working with placental perfusion methods calls for a great expenditure of technical details and time. A lot of physiological, biochemical, physical, and mathematical elements are involved and, therefore, for a structured group to follow this field over the years requires much effort, and not the least, tinancial. A parallel situation can be noted for journals. Placenta, a periodical published for the first time in 1980, and the American Journal of Obstetrics and Gynecology, which has the longest tradition and continuity in this field, both present contents closely related to the field of research. A peculiar position is occupied by Acta Endocrinologica, which published 22 papers, reflecting the fact that the most frequently examined substances in these reports were hormones. The techniques mainly used in these studies were perfusion on pregnant women and the human fetoplacental unit. Until 1973, 17 of the 22 papers were published by the same author. Both the author and the techniques subsequently vanished and this journal published only three more papers concerning studies using placental perfusion. A recurrent question was how closely the in vitro perfused placenta resembles the intrauterine placenta and how the performance of the in vitro preparations could be affected by traumatization of the structure and metabolic integrity following delivery.
OMARINIETAL. PLACENTAL
PERFUSION
There is ample evidence that despite all abuse the perfused placenta .retains many complex functions (e.g., active and energy-dependent transport of amino acids across the placental barrier by a stereospecific transport system that establishes a gradient towards the fetus; facilitated transport of o-glucose, which is three times higher than the diffusion rate of the L-isomer; and production of several placental peptides) (Dancis, 1985). A main consideration in this context is that the mother and fetus continuously modify placental “perfusates” in vivo. By removing the mother and fetus from the system, the perfusionist achieves a desirable simplicity and control of experimental conditions, so the confounding effects of maternal and fetal metabolism are avoided, and placental transfer and metabolism can be studied in isolation, Another recurrent question was whether the human or animal placenta is to be preferred, but there appears to be no clear answer. For practical, ethical, and legal reasons nowadays this question is closely related to the choice of using an in vitro or in vivo approach. There are some advantages to the in vivo approach, but such experiments are limited by concerns for the safety of mother and infant. These advantages are physiological, for example, original blood instead of highly artificial perfusates, no structural damage following separation from the uterine wall, and no anoxic period. Probably these were reasons why most researchers doing animal experiments prefer in vivo approaches and this also explains why the in situ animal perfusion technique has been more or less constant throughout the years. The fact that the guinea pig was the most widely selected species surely reflects the similarity of the hemochorial placenta to the human one, and the same reason is valid for the rabbit. An exceptional position is held by the sheep because most experimental work on placental hemodynamics has been done on this animal. Although there are fewer parallels between the epitheliochorial placenta and the human hemochorial one, in the sheep it is possible to perform intricate surgery and take fetal blood samples large enough for chemical analysis. It emerges clearly, however, that there is no ideal experimental animal for intrauterine surgery whose placenta is functionally and completely similar to the human placenta and fetuses are large enough to permit repeated sampling. It must also be borne in mind that the placenta is the mammalian organ that shows the greatest variations in structure among species (Faber and Thornburg, 1983). Differences in function must, therefore, also be assumed, and several important ones have already been documented. This statement is obviously reflected in the finding that in vitro perfusion of human placental
65
tissues has risen in the last year, whereas animal observations are more or less constant. Although attempts have been made to perfuse the whole placenta, methodological problems are such that nowadays this approach is nearly forgotten. Today many researchers follow the original protocol developed by Panigel (1968) or a modified version, perfusing a single cotyledon of a human placenta, with the maternal surface uppermost. An English group has developed a technique to perfuse the placenta with the maternal surface downwards, which seems to shorten the time between delivery and reperfusion (a crucial point in this technique) and avoids pooling maternal side effluents. Because these two approaches have never been compared in the same laboratory as far as we know, no distinct advantages of either one or the other version have been documented. With regard to the different topics that have been investigated with placental perfusion, most experiments were concerned with the transfer of substances across the placental barrier. Most of these investigated carbohydrates (particularly glucose as one of the most important fuels for the fetus), almost all known amino acids, and fatty acids. Many set out not only to answer the question whether these substances cross the placental barrier, but also to assess quantitative parameters. Metabolic reactions of the different substances constitute another main topic, and as a result of these experiments it has been demonstrated, for example, that the placenta is a metabolically active organ that has its own requirements because it consumes appreciable amounts of glucose and oxygen. Other experiments investigated the transfer of drugs, and the results are particularly valuable nowadays as almost all pregnant women receive some drug therapy. The fact that pharmacological perfusion experiments have increased in the last years reflects the increased sensitivity towards problems related to drug use in pregnancy. Some methodological problems emerged in reviewing all these articles. Despite the impressive amount of data, it remains very difficult for anyone to compare the results of studies with the same declared aim (i.e., studies on the same compounds, on similar physiological variables), because many papers fail to report some basic information. Kelmann and van Kreel (1987) proposed a few general criteria useful for judging the viability of placental perfusions in animals. Extending these criteria to perfusions of human placentas would provide a valuable basis for comparisons. In most of the papers (with the exception of declared “methodological” articles), information on the perfusion equipment were scant, and often even the indications for “open” or “closed” circulation were missing.
66
JPM Vol. 28, No. 2 September 1992:61-66
We again suggest that efforts should be made to report the main information to permit a comparisons of results. Examples are 1) measurements of maternal and fetal flows with related pressures, which could help one understand and minimize artifacts due to changes in permeability during perfusion; 2) possible maternal flow to the placenta in in vivo experiments; 3) if a drug or chemical agent was used during perfusion (anesthetics, heparin, antibiotics, antifoam, etc.); and 4) the composition of the perfusion medium and the pH measurements of flow in and out of the placenta (especially for pharmacological studies). It would also be useful if the variables-listed as a “perfusion index” (e.g., antipyrine, creatinine, and inulin transfer, oxygen balante)-were reported using the same international unit in different studies, as sometimes it is impossible to convert one unit to another. If the common intent is to provide a valid method to study more complex questions (for instance, effects of pharmacological substances on placental behavior) and, in general, study how compounds reach the fetus and their effects on the physiology of the placenta (and, therefore, on pregnancy), then the points briefly mentioned above constitute the minimum basis to enable other people working in the same field to use these studies as a constructive source of information. Even though an impressive amount of work has been done with these techniques, the approach is still more piecemeal than systematic, and, therefore, the full potential of this approach in biomedical research is far from being fully exploited.
of glucose, 800-806.
inulin and antipyrine.
Am .I Obstet Gynecol
146:
Carstensen M, Leichtweiss HP, Molsen G, Schroder H (1977) Evidence for a specific transport of o-hexoses across the human term placenta in vitro. Arch Gynaek 2221187-196. Contractor SF, Stannard PJ (1983) The use of AIB transport to assess the suitability of a system of human placental perfusion for drug transport studies. Placenta 4: 19-30. Dancis J (1985) Why perfuse the human placenta. In In Vitro Perfusion of Human Placental Tissue (vol 13 of Contribution to Gynecology and Obstetrics). Eds., H. Schneider and J. Dancis. Basel: Karger pp. l-4. Faber JJ, Thomburg KL (1983) Structural features of placental exchange. In Placental Physiology: Structure and Function ofFetomaternal Exchange. Eds., JJ Faber and KL Thornburg. New York: Raven Press, Inc, pp. l-32. Hamshaw-Thomas A, Reynolds F (1985) Placental transfer of bupivacaine, pethidine and lignocaine in the rabbit. Effect of umbilical flow rate and protein content. Br J Obstet Gynecol 92:706713.
Kelman BJ, van Kreel BK (1987) Criteria for standardizing and judging viability placental perfusions in animals. Trophoblast Research 2:515-522.
Mihaly GW, Morgan DJ (1984) Placental drug transfer: Effects of gestational age and species. Pharamcol Ther 23:253-266. Miller RK (1986) Techniques for the study of placental function. In Drug and Chemical Action in Pregnancy: Pharmacologic and Toxicologic Principles. Eds., S Fabro and AR Scialli. New York:
Marcel Dekker, Inc, pp. 128-135. Panigel M (1968) Placental perfusion. In Fetal Homeostasis, New York: Appleton Century Crofts, pp. 15-25.
vol 4.
Penfold P, Drury L, Simmonds R, Hytten FE (1981) Studies of a single placental cotyledon in vitro. I. The preparation and its viability. Placenta 2:149-154. Reynolds ML, Young M (1971) The transfer of free a-amino nitrogen across the placental membrane in the guinea-pig. J Physiol214: (Land) 583-597.
The authors thank Dr. Wolfgang Reiber for a critical reading of the manuscript, Lucia August0 Casaroli for computer assistance, Judy Baggott for the editing, and Daniela Miglio for the secretarial help. The generous contribution of the Banca Popolare, Milan, Italy, is also gratefully acknowledged.
Schneider H (1991) Placental transport function. Reprod Fertil Dev 3:345-353.
Schneider H, Huch A (1985) Dual in vitro perfusion of an isolated lobe of human placenta: Method and instrumentation. In In vitro Perfusion of Human Placental Tissue (vol 13 of Contribution to Gynecology and Obstetrics). Eds., H Schneider and J Dancis. Basel: Karger pp. 40-45.
References
Shaw AJ, Mughal MZ, Mohammed T, Maresh MJA, Sibley CP (1990) Evidence for active matemofetal transfer of magnesium across the in situ perfused rat placenta. Pediatr Res17(6):
Brandes JM, Tavolini N, Potter BJ, Sarkozi L, Shepard MD, Berk PD (1983) A new recycling technique for human placental cotyledon perfusion: Application to studies of the fetomaternal transfer
Stulc J, Stulcova B (1986) Transport of calcium by the placenta of the rat. J Physiol (Lond) 371:1-16.
622-625.
