- Email: [email protected]
Investigation of the acid-base balance of coelomic and amniotic fluids in early human pregnancy
Investigation of the acid-base balance of coelomic and amniotic fluids in early human pregnancy Eric Jauniaux, MD, PhD," Davor Jurkovic, MD, PhD," Beatrice Gulbis, MD, PhD,b William P. Collins, PhD," Jamal Zaidi, MD," and Stuart Campbell, MD" London, Unzted Kmgdom, and Brussels, Belgium OBJECTIVE: The aim of our study was to analyze and compare the acid-base status of coelomic and amniotic fluids in early human pregnancy. STUDY DESIGN: Coelomic fluid, amniotic fluid, and maternal serum were obtained from 55 normal pregnancies between 7 and 14 weeks of gestation and assayed for pH, PC02' bicarbonate, chloride, phosphate, total protein, glucose, and lactate concentrations. RESULTS: The coelomic fluid had a significantly lower pH and base excess and contained significantly lower glucose, total protein, and bicarbonate concentrations and higher Pco2, lactate, and phosphate levels than did maternal serum. In the amniotic fluid significantly higher pH, base excess, and lactate and bicarbonate levels and lower concentrations of chloride and phosphate were found compared with the coelomic fluid. Little variation was observed in the coelomic and amniotic fluid composition before 11 weeks of gestation. A significant increase in Pc0 2 , total protein, and chloride levels and a decrease in pH, base excess, and bicarbonate concentrations were observed in the amniotic fluid between 11 and 14 weeks compared with 7 to 10 weeks. CONCLUSIONS: The coelomic fluid composition was indicative of an anaerobic metabolic aCidosis probably related to the accumulation of acid compounds from placental metabolism. During the same period of gestation the composition of the amniotic fluid demonstrated a metabolic alkalosis that probably arises from the accumulation of basic substances through the unkeratinized embryonic skin and from the metabolism of organic anions in the embryonic tissues. The changes in the amniotic composition and acid-base balance at the end of the first trimester may correspond to the switch from the mesonephros to the metanephros and result from the excretion in the amniotic fluid of acid metabolic fetal bioproducts through the maturing kidneys. (AM J OesTET GYNECOL 1994;170:1365-9.)
Key words: Early human pregnancy, coelomic fluid, amniotic fluid, embryo, fetus
The transport of nutrients, gases, hormones, and immunologic factors from maternal to fetal blood, and vice versa, has been extensively investigated during the second half of human gestation. I. 2 When the definitive placenta is formed, bioproducts present in the maternal and fetal blood are passively or actively transported, depending on their biochemical characteristics, through the placental villous barrier or are metabolized within the placental tissue. 1 In addition to this major route, bioproducts either excreted in the amniotic fluid
From the Earl.v Human Development Unzt, Department of Obstetncs and Gynaecology, King's College School of MediCine and DenttStry, Unzverstty of London, a and the Department of Clinical ChemtStry, AcademiC Hospital Erasme, Free UnIVerstty of Brussels! Supported by a grant of the Fondation Universitatre David & Alice Van Buren. ReceIVed for publicatwn September 22, 1993; revised November 23, 1993; accepted December 9, 1993. Reprint requests: Eric Jauniaux, MD, PhD, Department of Obstetncs and Gynaecology, King's College Hospital, Denmark Hill, London, Unzted Kingdom SE5 8RX. Copynght © 1994 by Mosby-Year Book, Inc. 0002-9378/94 $3.00 + 0 6/1/53553
through the fetal urinary, respiratory, and digestive tracts or present in the maternal circulation may also be transferred through the amniotic epithelium and the decidual layers of the free placental membrane." The first-trimester gestational sac is structurally and functionally different from the second- and third-trimester sac. Up to 12 weeks of gestation the amniotic cavity containing the developing embryo is separated from the placenta by the exocoelomic or chorionic cavity containing the secondary yolk sac. 4 . 5 Furthermore, up to 10 weeks of gestation the whole external surface of the gestational sac is covered by placental tissue. Major anatomic changes are observed between the end of the first trimester and the beginning of the second trimester, which consist mainly in the disappearance of the exocoelomic cavity, the secondary yolk sac, and two thirds of the primitive placental mass! The aim of this study was to investigate the acid-base balance of coelomic and amniotic fluids and to evaluate the possible role of changes in the gestational sac anatomy and of fetal organ development on the embryonic fluids biochemistry. 1365
1366 Jauniaux et al.
May 1994 Am J Obstet Gynecol
Patients and methods
Coelomic fluid, amniotic fluid, and maternal serum were obtained from 55 healthy women with apparently normal pregnancies who were undergoing termination for psychosocial reasons between 7 and 14 weeks of gestation. The mean maternal age was 21 years, and the mean parity was 1.54. In all cases the fetal heart beat was confirmed, and the crown-rump length was within the normal range for gestational age. Written informed consent was obtained from each patient before the surgical procedure. Coelomic and amniotic fluids were retrieved by transvaginal puncture under ultrasonographic guidance, as described previously.6 When the two compartments were dearly visualized, exocoelomic fluid was aspirated by gentle aspiration with a 20-gauge needle. Subsequently a new 20-gauge needle was reintroduced through the guide and advanced into the amniotic cavity to aspirate amniotic fluid. All the fluid was aspirated. The first 0.2 ml of each fluid was discarded to diminish the risk of contamination. In all cases 10 mlof maternal venous blood was taken from a forearm vein at the time of pregnancy termination. For pH, Pco 2 , bicarbonate and base excess measurements 1 ml of each fluid and of maternal venous blood was collected directly into preheparinized plastic syringes and immediately analyzed by a Radiometer ABL automatic analyzer (Radiometer, Copenhagen). Aliquots of coelomic and amniotic fluid and blood serum were also stored at -70 C. Samples used for the measurements of glucose and lactate were stored in sodium fluoride tubes. All samples were assayed at the same time for glucose, lactate, total protein, chloride, and phosphate, and all measurements were performed by a Hitachi 717 automatic analyzer (Tokyo) with commercially available kits (Boehringer-Mannheim, Mannheim, Germany). A dye-binding method (Sopachem 003.0309.02, Sopar Biochem, Brussels) was used to measure total protein concentration by spectrophotometry. Glucose and lactate levels were determined by enzymatic methods. A selective ion electrode was used for the estimation of chloride level. Inorganic phosphate measurements involved the use of phosphomolybdate. Linear regression equations were calculated by the least-squares method and their slope tested for significance by the F ratio test. A Kruskal-Wallis rank test was performed to compare biologic values in the different compartments and in amniotic fluid at two different periods of gestation. A p :5 0.05 was considered to indicate the presence of significance. 0
Results
Coelomic fluid was aspirated in 33 pregnancies between 7.3 and 11.1 weeks of gestation, and amniotic
fluid was obtained in 47 pregnancies between 7.9 and 13.9 weeks. Matched samples of coelomic and amniotic fluids and maternal peripheral blood were collected from 24 patients between 7.9 and 11.1 weeks. The volume of fluid obtained from each of the embryonic cavities varied between 2 and 7 m!. The coelomic fluid was always yellow and more viscous than the amniotic fluid. The mean and distribution of values from the various assays in the different compartments are compared in Table I. Significantly higher Pco 2 , lactate, and phosphate and lower pH, base excess, glucose, bicarbonate, and total protein were found in the coelomic fluid compared with maternal serum. No difference was found for chloride level between coelomic fluid and maternal serum. Higher pH, base excess, lactate, and bicarbonate levels and lower concentrations of chloride and phosphate were found in the amniotic fluid compared with the coelomic fluid. No difference was found in Pco 2 and glucose concentration between the two embryonic fluids. Significant correlations were found between gestational age and coelomic fluid lactate and total protein concentrations and between gestational age and amniotic pH and base excess and amniotic fluid levels of bicarbonate, total protein, and chloride (Table II). Fig. 1 shows the mean and SD value for pH in the coelomic and amniotic fluids by weeks of gestation. Important changes were observed around 11 weeks in the amniotic fluid when the pH decreased. The mean levels of the different biochemical parameters measured in the amniotic fluid between 7 and 10 and 11 and 14 weeks' gestation are compared in Table III. Significantly higher Pc0 2 , total protein, and chloride levels were found between 7 and 10 weeks compared with 11 to 14 weeks, whereas the pH, base excess, and bicarbonate concentrations were significantly lower. No difference was found between the two gestational periods for the levels of glucose and lactate. Pc0 2 and glucose levels in coelomic and amniotic fluids were significantly positively correlated (Table II). Significant (p < 0.05) negative linear correlations were also found between gestational age and maternal serum levels of glucose, bicarbonate, and total protein. No correlation was found for the different variables between coelomic fluid and maternal serum. In amniotic fluid samples collected before 11 weeks only pH and bicarbonate level were significantly (p < 0.01) correlated with gestational age. Comment
The result of this investigation demonstrate that at the same period of pregnancy the acid-base balance regulation and the biochemical composition of the coelomic and amniotic fluids are different and that
Jauniaux et al.
Volume 170, Number 5, Part I Am J Obstet Gynecol
1367
Table I. Comparison of the mean (± SD) values for pH, Pco 2 , base excess, glucose, lactate, total protein, and electrolytes in 24 matched samples of maternal serum, coelomic fluid, and amniotic fluid
Maternal serum
Vanables
7.38 43 -2.6 3.4 0.3 22 67.9 105 1.01
pH Pco 2 (mm Hg) Base excess (mmol/L) Glucose (mmol/L) Lactate (mmol/L) Bicarbonate (mmol/L) Total protein (gm/L) Chloride (mmol/L) Phosphate (mmol/L)
± ± ± ± ± ± ± ± ±
Coelomic fluid
0.03 6 1.9 0.9 0.2 3 5.3 2 0.19
7.18 56 -7.8 2.7 0.6 18 3.7 105 2.1
Amnwtic fluid
0.09 11 3.7 0.7 0.2 2 1.1 6 0.26
± ± ± ± ± ± ± ± ±
7.42 53 10.2 2.8 0.9 38 0.1 92 0.71
± ± ± ± ± ± ± ± ±
0.09 10 6.8 0.5 0.2 6 0.04 8 0.32
Maternal serum vs coelomic fluid (significance)
Codomlc fluid vs amnwtlc flUid (significance)
P<
P < 0.0001 p < 0.001
P<
P<
0.001
P < 0.005 P < 0.005 p < 0.0001
P< P< P< P< P<
P<
0.0001 NS P < 0.0001
0.0001 NS 0.0001 NS 0.005 0.0001 0.0001 0.05 0.0001
NS, Not significant.
Table II. Correlations between gestational age and levels of different biochemical variables in coelomic and amniotic fluids and between both fluid cavities Changes with gestatwnal age Coelomic fluzd (n = 33)
I
Variables
r
pH Pco 2 (mm Hg) Base excess (mmol/L) Glucose (mmol/L) Lactate (mmol/L) Bicarbonate (mmol/L) Total protein (gm/L) Chloride (mmol/L)
-0.03 -0.17 -0.12
0.1 0.7 0.5
-0.15 0.47 -0.15
0.7 8.2 0.6
0.43
6.5
-0.25
1.9
F
I Significance
AmnIOtic fluid (n = 47)
r
I
F
I Significance
Coelomic fluid liS amniotic fluid (n = 24) r
I
F
NS NS NS
-0.77 -0.25 -0.88
64 2.9 148
P<
0.0001 NS P < 0.0001
0.02 0.39 0.24
0.1 8.2 3.3
P=
NS 0.008 NS
0.14 0.03 -0.87
0.9 0.1 145
NS NS P < 0.0001
0.63 0.11 0.17
14.3 0.3 0.6
P=
0.02
NS
I
Significance
P= P=
NS 0.006 NS 0.001 NS NS
0.73
52
P<
0.0001
0.35
3.2
NS
0.59
24
P<
0.0001
0.22
1.2
NS
NS, Not significant.
important changes occur in the amniotic fluid composition at the end of the first trimester. Until recently amniotic fluid was the only relatively accessible fetal milieu in early pregnancy. Amniotic fluid analyses have been performed from the third month of gestation onward and have demonstrated important variation in gas tension, acid-base status, and biochemical composition with gestational age and differences compared with maternal blood. 7 • R The amniotic fluid biochemistry has been occasionally investigated in the first trimester,"' 10 but the changes with advancing gestation were never evaluated at the time of the shift from the embryonic to the fetal period. Because of technical difficulties in accurately recognizing the different anatomic structure of the early gestational sac in utero, it is likely that some of the rare samples obtained in these studies before 12 weeks of gestation were a mixture of coelomic and amniotic fluids. The advent of high-resolution transvaginal ultrasonography
transducers has enabled a detailed morphologic assessment of the early gestational sac. In particular, the membrane separating the exocoelomic and amniotic cavities can now be clearly identified in utero, and coelomic fluid can be selectively aspirated between 6 and 12 weeks of gestation in sufficient amount for biochemical analysis 7 and genetic investigations. I I In vivo and in vitro experiments have recently demonstrated that there is no real continuous maternal blood circulation in the intervillous space before 12 weeks of gestation. 12. 13 During the first 3 months of gestation the villi are bathed in a clear fluid probably made of filtered maternal plasma and uterine gland secretions. We have recently measured in vivo the oxygen level in early human pregnancy by inserting polarographic electrodes inside the placenta and the endometrium. H Between 8 and 10 weeks of gestation significantly lower P0 2 levels were found in the placenta compared with the endometrium, whereas placental
1368 Jauniaux et al.
May 1994 Am J Obstet Gynecol
7.6
:r
co.
~
7.6
7.5
:rQ.
7.4
~
·s
E 0
7.2
..
0
7.1
~
·5
0: 7.3
i;: 7.3
U
a; 0
7.5 7.4
U
07.2
"E
7.1
6.9
6.9 6
8
10
11
+---,---.,.--...---r--...,---,-8
Gestation (weeks)
10
11
12
13
Gestation (weeks)
Fig. 1. Mean value (±SD) by weeks of gestation for coelomic fluid pH (left) and amniotic fluid pH (TIght).
Table III. Comparison of mean value (± SD) of different biochemical variables obtained in amniotic fluid between 7 and 10 weeks and 11 and 13 weeks of gestation Variables
pH Pco 2 (mm Hg) Base excess (mmoI!L) Glucose (mmoI!L) Lactate (mmoI!L) Bicarbonate (mmoI!L) Total protein (gm/L) Chloride (mmoI!L)
7-10 wk (n = 23)
7.45 50 12.4 2.7 1.0 39 0.05 90
± 0.06 ± 7 ± 3.8
± 0.6
± 0.2
± 5.6 ± 0.04 ± 9.3
11-14 wk (n = 24)
7.23 55 -5.1 2.8 0.9
± 0.09 ± 10 ± 6.2 ± 0.5 ± 0.4 21.1 ± 5.6 0.36 ± 0.4 102 ± 8.2
Rank test (sIgnificance)
p < 0.0001
P< P< P<
0.05 0.0001
NS NS
0.0001
P < 0.0001
P<
0.0001
NS, Not Significant.
and endometrial P0 2 levels were similar after 12 weeks. The increase of the placental P0 2 at the end of the first trimester has been related to the establishment of the maternal circulation in the intervillous chamber, which occurs during that period of gestation. The results of these investigations suggest that the embryo develops in an environment poorer in oxygen than that of the fetus and is separated from the maternal circulation by a more complex barrier. As a consequence of respiratory alkalosis of pregnancy, maternal renal excretion of bicarbonate secondarily increases, and the overall maternal blood pH remains relatively unchanged. I5 Except for the total protein and lactate levels, we found no significant variation in the coelomic fluid biochemical composition between 7 and 11 weeks of gestation and no correlation for the different variables between coelomic fluid and maternal serum, suggesting that the coelomic fluid acid-base regulation is not directly influenced by maternal physiologic changes. During that period the coelomic fluid had a lower pH, base excess, and bicarbonate level than did maternal venous blood and contained higher levels of Pc0 2 , lactate, and phosphate and lower levels of protein than did the maternal serum (Table I). These findings are consistent with a metabolic anaerobic acidosis and suggest the accumulation of acid bioproducts from the placental metabolism in the exocoelomic cavity.
Specific placental proteins such as human chorionic gonadotropin are found in higher concentrations in the coelomic fluid than in maternal serum, suggesting the presence of a preferential transfer of placental bioproducts toward the exocoelomic cavity. Jfi The significant drop in glucose level from maternal serum to the coelomic fluid probably reflects the important consumption of this substance by the placental tissue (Table I). More than half of the carbon dioxide produced by the gestational sac metabolism is removed in this form through the placenta. I7 Carbonic anhydrase activity has been demonstrated in the human villous tissue as early as the sixth week of gestation. 1" Experiments with animals have shown that there are at least two mechanisms for placental bicarbonate transfer (i.e., by conversion to carbon dioxide, which then diffuses across the villous barrier or by using an anion exchange system involving chloride and lactate ions I7 ). The latter mechanism could explain the high lactate concentration found in the coelomic fluid, which is known to be a major substrate for the energy required by the growing embryo or fetus. 2 High concentrations of phosphate in the coelomic fluid (Table I) suggest that this molecule serves as a buffer for the coelomic fluid acid-base balance. By contrast, the role of proteins is probably limited, because their level in the coelomic fluid were about 18 times lower than those of matched samples of maternal serum. These data suggest that the coelomic
Jauniaux et al.
Volume 170, Number 5, Part I Am J Obstet Gynecol
fluid composition is mainly influenced by the biologic activities of the placenta, which forms a continuous barrier to the maternal circulation and is the dominant embryonic organ at this time of gestation. During the second and third trimester the pool of amniotic fluid is subject to a constant turnover, being added to by fetal lung fluid and urine and being removed by fetal swallowing.' When the definitive placenta has formed, movements of water and electrolyte also take place across the free placental membranes. 3 The amniotic fluid collected before 11 weeks of gestation had a higher pH and base excess, higher levels of lactate and bicarbonate, and lower levels of total protein, phosphate, and chloride than did the coelomic fluid collected during the same period of gestation (Table I). Among these biologic parameters only pH and bicarbonate levels varied between 7 and 11 weeks of gestation. The metabolic alkalosis of the amniotic fluid observed in the first trimester probably arises from the accumulation of bicarbonate and the increased metabolism of organic anions such as lactate in the embryonic tissue. The embryonic skin, which only becomes keratinized during the second trimester,' is probably the major source of amniotic fluid in early pregnancy. The development of the ruinary and respiratory tracts and the rupture of the cloacal membrane during the eighth week of gestation must also influence the amniotic fluid composition from the third month of pregnancy. In particular, the mesonephros or temporary kidney that develops between embryonic stages 10 and 14 is capable of producing urine from embryonic state 17, which corresponds to 7.5 weeks after the last menstruation. 19 The metanephros or permanent kidney starts to form at embryonic stage 14 and produces urine around 10 weeks of gestation. 19 Significant changes were observed in the amniotic fluid composition at the end of the first trimester characterized by a decrease of pH (Fig. 1), base excess, and bicarbonate level and an increase of Pco 2 and chloride levels (Tables II and III). These changes of the amniotic composition continue during the second and third trimesters 8 - lo • 19 and probably reflect the increasing contribution of the fetal metaboli'sm to the amniotic fluid composition. It is of interest that except for the total protein level, which remained lower, the mean value of the other amniotic biologic parameters obtained after 11 weeks of gestation became similar to those found in the coelomic fluid before 11 weeks (Table I). We suggest that the changes of the amniotic fluid composition observed at the end of the first trimester correspond to the switch from the mesonephros to the metanephros and that from this stage of gestation acid bioproducts produced in increasing amounts by the metabolism of the rapidly growing fetus are mainly eliminated in the amniotic cavity by the immature kidneys.
1369
We thank the King's Health Care Authority and the staff of the Day Surgery Centre and Labour ward for their constant support. REFERENCES I. Sibley CP, Boyd RDH. Mechanism of transfer across the human placenta. In: Polin RA, Fox WW. eds. Fetal and neonatal physiology. Philadelphia: WB Saunders, 1992:6274. 2. Burd LJ, Jones MD, Simmonds MA, Makowski EL, Meschia G, Battaglia FC. Placental production and foetal utilization of lactate and pyruvate. Kature 1975;254: 710-2. 3. MacCarthy T, Saunders P. The origin and Circulation of the amniotic fluid. In: Fairweather DVI. Eskes TKAB, eds. Amniotic fluid: research and clinical application. 2nd ed. Amsterdam: Excerpta Medica, 1978:1-18. 4. BoydJD, Hamilton WJ. The human placenta. Cambridge: Heffer, 1970. 5. Jauniaux E, Burton GJ, Jones CP]. Early human placental morphology. In: Barnea E, Hustin J, Jauniaux E, eds. The first twelve weeks of gestation. Heidelberg: Springer-Verlag, 1992:45-64. 6. Jauniaux E, Jurkovic D, Gulbis B, Gervy C, Ooms HA, Campbell S. Biochemical composition of exocoeiomlC fluid in early human pregnancy. Obstet GynecoI1991;78: 1124-8. 7. Lind T. The biochemistry of amniotic fluid. In: Fairweather DVI, Eskes TKAB. Amniotic fluid: research and clinical application. 2nd ed. Amsterdam: Excerpta Medica, 1978:59-80. 8. Economides DI, Johnson MA, MacKenzie IZ. Does amniotic fluid analysis reflect acid-base balance in fetal blood. AM J OBSTET GYNECOL 1992;166:970-3. 9. Johnell HE, Nilsson BA. Oxygen tension, acid-base status and electrolytes in human amniotic fluid. Acta Obstet Gynecol Scand 1971 ;50: 183-92. 10. Sinha R, Carlton M. The volume and composition of amnotic fluid in early pregnancy. J Obstet Gynaecol Br Commonw 1970;77:211-4. II. Jurkovic D, Jauniaux E, Campbell S, Pandya P, Cardy DL, Nicolaides KH. Coelocentesis: a new technique for early prenatal diagnosis. Lancet 1993;341: 1623-4. 12. HustinJ, SchaapsJP. Echographic and anatomic studies of the maternotrophoblastic border during the first trimester of pregnancy. AM J OBSH:T GYNECOL 1987; 157: I 62-8. 13. Hustin.J, Schaaps .JP, Lambotte R. Anatomical studies of the uteroplacental vascularization in the first trimester of pregnancy. Trophoblast Res 1988;3:49-60. 14. Rodesch F, Simon P, Donner C, Jauniaux E. Oxygen measurements in the maternotrophoblastic border during early pregnancy. Obstet Gynecol 1992;80:283-285. 15. Lim VS, Katz AI, Lindheimer MD. ACId-base regulation in pregnancy. AmJ PhysioI1976;231:1764-70. 16. Jauniaux E, Gulbis B, Jurkovic D, Schaaps JP, Campbell S, Meuris S. Protein and steroid level, in embryonic cavities of early human pregnancy. Hum Reprod 1993;8:782-7. 17. Hatano H, Leichtweij3 HP, Schroder H. Uptake of bicarbonate/C0 2 in the isolated guinea pig placenta. Placenta 1989; 10:212-21. 18. Aliakbar S, Brown PR, Jauniaux E, Bidwell DE, Kicolaides KH. Measurements of carbonic anhydrase Isoenzyme, in early human placental tissues. Biochem Soc Tram 1990; 18:670. 19. O'Rahilly R, Muller F. Human embryology and teratology. New York: Whiley-Liss, 1992.
Key words: Early human pregnancy, coelomic fluid, amniotic fluid, embryo, fetus
The transport of nutrients, gases, hormones, and immunologic factors from maternal to fetal blood, and vice versa, has been extensively investigated during the second half of human gestation. I. 2 When the definitive placenta is formed, bioproducts present in the maternal and fetal blood are passively or actively transported, depending on their biochemical characteristics, through the placental villous barrier or are metabolized within the placental tissue. 1 In addition to this major route, bioproducts either excreted in the amniotic fluid
From the Earl.v Human Development Unzt, Department of Obstetncs and Gynaecology, King's College School of MediCine and DenttStry, Unzverstty of London, a and the Department of Clinical ChemtStry, AcademiC Hospital Erasme, Free UnIVerstty of Brussels! Supported by a grant of the Fondation Universitatre David & Alice Van Buren. ReceIVed for publicatwn September 22, 1993; revised November 23, 1993; accepted December 9, 1993. Reprint requests: Eric Jauniaux, MD, PhD, Department of Obstetncs and Gynaecology, King's College Hospital, Denmark Hill, London, Unzted Kingdom SE5 8RX. Copynght © 1994 by Mosby-Year Book, Inc. 0002-9378/94 $3.00 + 0 6/1/53553
through the fetal urinary, respiratory, and digestive tracts or present in the maternal circulation may also be transferred through the amniotic epithelium and the decidual layers of the free placental membrane." The first-trimester gestational sac is structurally and functionally different from the second- and third-trimester sac. Up to 12 weeks of gestation the amniotic cavity containing the developing embryo is separated from the placenta by the exocoelomic or chorionic cavity containing the secondary yolk sac. 4 . 5 Furthermore, up to 10 weeks of gestation the whole external surface of the gestational sac is covered by placental tissue. Major anatomic changes are observed between the end of the first trimester and the beginning of the second trimester, which consist mainly in the disappearance of the exocoelomic cavity, the secondary yolk sac, and two thirds of the primitive placental mass! The aim of this study was to investigate the acid-base balance of coelomic and amniotic fluids and to evaluate the possible role of changes in the gestational sac anatomy and of fetal organ development on the embryonic fluids biochemistry. 1365
1366 Jauniaux et al.
May 1994 Am J Obstet Gynecol
Patients and methods
Coelomic fluid, amniotic fluid, and maternal serum were obtained from 55 healthy women with apparently normal pregnancies who were undergoing termination for psychosocial reasons between 7 and 14 weeks of gestation. The mean maternal age was 21 years, and the mean parity was 1.54. In all cases the fetal heart beat was confirmed, and the crown-rump length was within the normal range for gestational age. Written informed consent was obtained from each patient before the surgical procedure. Coelomic and amniotic fluids were retrieved by transvaginal puncture under ultrasonographic guidance, as described previously.6 When the two compartments were dearly visualized, exocoelomic fluid was aspirated by gentle aspiration with a 20-gauge needle. Subsequently a new 20-gauge needle was reintroduced through the guide and advanced into the amniotic cavity to aspirate amniotic fluid. All the fluid was aspirated. The first 0.2 ml of each fluid was discarded to diminish the risk of contamination. In all cases 10 mlof maternal venous blood was taken from a forearm vein at the time of pregnancy termination. For pH, Pco 2 , bicarbonate and base excess measurements 1 ml of each fluid and of maternal venous blood was collected directly into preheparinized plastic syringes and immediately analyzed by a Radiometer ABL automatic analyzer (Radiometer, Copenhagen). Aliquots of coelomic and amniotic fluid and blood serum were also stored at -70 C. Samples used for the measurements of glucose and lactate were stored in sodium fluoride tubes. All samples were assayed at the same time for glucose, lactate, total protein, chloride, and phosphate, and all measurements were performed by a Hitachi 717 automatic analyzer (Tokyo) with commercially available kits (Boehringer-Mannheim, Mannheim, Germany). A dye-binding method (Sopachem 003.0309.02, Sopar Biochem, Brussels) was used to measure total protein concentration by spectrophotometry. Glucose and lactate levels were determined by enzymatic methods. A selective ion electrode was used for the estimation of chloride level. Inorganic phosphate measurements involved the use of phosphomolybdate. Linear regression equations were calculated by the least-squares method and their slope tested for significance by the F ratio test. A Kruskal-Wallis rank test was performed to compare biologic values in the different compartments and in amniotic fluid at two different periods of gestation. A p :5 0.05 was considered to indicate the presence of significance. 0
Results
Coelomic fluid was aspirated in 33 pregnancies between 7.3 and 11.1 weeks of gestation, and amniotic
fluid was obtained in 47 pregnancies between 7.9 and 13.9 weeks. Matched samples of coelomic and amniotic fluids and maternal peripheral blood were collected from 24 patients between 7.9 and 11.1 weeks. The volume of fluid obtained from each of the embryonic cavities varied between 2 and 7 m!. The coelomic fluid was always yellow and more viscous than the amniotic fluid. The mean and distribution of values from the various assays in the different compartments are compared in Table I. Significantly higher Pco 2 , lactate, and phosphate and lower pH, base excess, glucose, bicarbonate, and total protein were found in the coelomic fluid compared with maternal serum. No difference was found for chloride level between coelomic fluid and maternal serum. Higher pH, base excess, lactate, and bicarbonate levels and lower concentrations of chloride and phosphate were found in the amniotic fluid compared with the coelomic fluid. No difference was found in Pco 2 and glucose concentration between the two embryonic fluids. Significant correlations were found between gestational age and coelomic fluid lactate and total protein concentrations and between gestational age and amniotic pH and base excess and amniotic fluid levels of bicarbonate, total protein, and chloride (Table II). Fig. 1 shows the mean and SD value for pH in the coelomic and amniotic fluids by weeks of gestation. Important changes were observed around 11 weeks in the amniotic fluid when the pH decreased. The mean levels of the different biochemical parameters measured in the amniotic fluid between 7 and 10 and 11 and 14 weeks' gestation are compared in Table III. Significantly higher Pc0 2 , total protein, and chloride levels were found between 7 and 10 weeks compared with 11 to 14 weeks, whereas the pH, base excess, and bicarbonate concentrations were significantly lower. No difference was found between the two gestational periods for the levels of glucose and lactate. Pc0 2 and glucose levels in coelomic and amniotic fluids were significantly positively correlated (Table II). Significant (p < 0.05) negative linear correlations were also found between gestational age and maternal serum levels of glucose, bicarbonate, and total protein. No correlation was found for the different variables between coelomic fluid and maternal serum. In amniotic fluid samples collected before 11 weeks only pH and bicarbonate level were significantly (p < 0.01) correlated with gestational age. Comment
The result of this investigation demonstrate that at the same period of pregnancy the acid-base balance regulation and the biochemical composition of the coelomic and amniotic fluids are different and that
Jauniaux et al.
Volume 170, Number 5, Part I Am J Obstet Gynecol
1367
Table I. Comparison of the mean (± SD) values for pH, Pco 2 , base excess, glucose, lactate, total protein, and electrolytes in 24 matched samples of maternal serum, coelomic fluid, and amniotic fluid
Maternal serum
Vanables
7.38 43 -2.6 3.4 0.3 22 67.9 105 1.01
pH Pco 2 (mm Hg) Base excess (mmol/L) Glucose (mmol/L) Lactate (mmol/L) Bicarbonate (mmol/L) Total protein (gm/L) Chloride (mmol/L) Phosphate (mmol/L)
± ± ± ± ± ± ± ± ±
Coelomic fluid
0.03 6 1.9 0.9 0.2 3 5.3 2 0.19
7.18 56 -7.8 2.7 0.6 18 3.7 105 2.1
Amnwtic fluid
0.09 11 3.7 0.7 0.2 2 1.1 6 0.26
± ± ± ± ± ± ± ± ±
7.42 53 10.2 2.8 0.9 38 0.1 92 0.71
± ± ± ± ± ± ± ± ±
0.09 10 6.8 0.5 0.2 6 0.04 8 0.32
Maternal serum vs coelomic fluid (significance)
Codomlc fluid vs amnwtlc flUid (significance)
P<
P < 0.0001 p < 0.001
P<
P<
0.001
P < 0.005 P < 0.005 p < 0.0001
P< P< P< P< P<
P<
0.0001 NS P < 0.0001
0.0001 NS 0.0001 NS 0.005 0.0001 0.0001 0.05 0.0001
NS, Not significant.
Table II. Correlations between gestational age and levels of different biochemical variables in coelomic and amniotic fluids and between both fluid cavities Changes with gestatwnal age Coelomic fluzd (n = 33)
I
Variables
r
pH Pco 2 (mm Hg) Base excess (mmol/L) Glucose (mmol/L) Lactate (mmol/L) Bicarbonate (mmol/L) Total protein (gm/L) Chloride (mmol/L)
-0.03 -0.17 -0.12
0.1 0.7 0.5
-0.15 0.47 -0.15
0.7 8.2 0.6
0.43
6.5
-0.25
1.9
F
I Significance
AmnIOtic fluid (n = 47)
r
I
F
I Significance
Coelomic fluid liS amniotic fluid (n = 24) r
I
F
NS NS NS
-0.77 -0.25 -0.88
64 2.9 148
P<
0.0001 NS P < 0.0001
0.02 0.39 0.24
0.1 8.2 3.3
P=
NS 0.008 NS
0.14 0.03 -0.87
0.9 0.1 145
NS NS P < 0.0001
0.63 0.11 0.17
14.3 0.3 0.6
P=
0.02
NS
I
Significance
P= P=
NS 0.006 NS 0.001 NS NS
0.73
52
P<
0.0001
0.35
3.2
NS
0.59
24
P<
0.0001
0.22
1.2
NS
NS, Not significant.
important changes occur in the amniotic fluid composition at the end of the first trimester. Until recently amniotic fluid was the only relatively accessible fetal milieu in early pregnancy. Amniotic fluid analyses have been performed from the third month of gestation onward and have demonstrated important variation in gas tension, acid-base status, and biochemical composition with gestational age and differences compared with maternal blood. 7 • R The amniotic fluid biochemistry has been occasionally investigated in the first trimester,"' 10 but the changes with advancing gestation were never evaluated at the time of the shift from the embryonic to the fetal period. Because of technical difficulties in accurately recognizing the different anatomic structure of the early gestational sac in utero, it is likely that some of the rare samples obtained in these studies before 12 weeks of gestation were a mixture of coelomic and amniotic fluids. The advent of high-resolution transvaginal ultrasonography
transducers has enabled a detailed morphologic assessment of the early gestational sac. In particular, the membrane separating the exocoelomic and amniotic cavities can now be clearly identified in utero, and coelomic fluid can be selectively aspirated between 6 and 12 weeks of gestation in sufficient amount for biochemical analysis 7 and genetic investigations. I I In vivo and in vitro experiments have recently demonstrated that there is no real continuous maternal blood circulation in the intervillous space before 12 weeks of gestation. 12. 13 During the first 3 months of gestation the villi are bathed in a clear fluid probably made of filtered maternal plasma and uterine gland secretions. We have recently measured in vivo the oxygen level in early human pregnancy by inserting polarographic electrodes inside the placenta and the endometrium. H Between 8 and 10 weeks of gestation significantly lower P0 2 levels were found in the placenta compared with the endometrium, whereas placental
1368 Jauniaux et al.
May 1994 Am J Obstet Gynecol
7.6
:r
co.
~
7.6
7.5
:rQ.
7.4
~
·s
E 0
7.2
..
0
7.1
~
·5
0: 7.3
i;: 7.3
U
a; 0
7.5 7.4
U
07.2
"E
7.1
6.9
6.9 6
8
10
11
+---,---.,.--...---r--...,---,-8
Gestation (weeks)
10
11
12
13
Gestation (weeks)
Fig. 1. Mean value (±SD) by weeks of gestation for coelomic fluid pH (left) and amniotic fluid pH (TIght).
Table III. Comparison of mean value (± SD) of different biochemical variables obtained in amniotic fluid between 7 and 10 weeks and 11 and 13 weeks of gestation Variables
pH Pco 2 (mm Hg) Base excess (mmoI!L) Glucose (mmoI!L) Lactate (mmoI!L) Bicarbonate (mmoI!L) Total protein (gm/L) Chloride (mmoI!L)
7-10 wk (n = 23)
7.45 50 12.4 2.7 1.0 39 0.05 90
± 0.06 ± 7 ± 3.8
± 0.6
± 0.2
± 5.6 ± 0.04 ± 9.3
11-14 wk (n = 24)
7.23 55 -5.1 2.8 0.9
± 0.09 ± 10 ± 6.2 ± 0.5 ± 0.4 21.1 ± 5.6 0.36 ± 0.4 102 ± 8.2
Rank test (sIgnificance)
p < 0.0001
P< P< P<
0.05 0.0001
NS NS
0.0001
P < 0.0001
P<
0.0001
NS, Not Significant.
and endometrial P0 2 levels were similar after 12 weeks. The increase of the placental P0 2 at the end of the first trimester has been related to the establishment of the maternal circulation in the intervillous chamber, which occurs during that period of gestation. The results of these investigations suggest that the embryo develops in an environment poorer in oxygen than that of the fetus and is separated from the maternal circulation by a more complex barrier. As a consequence of respiratory alkalosis of pregnancy, maternal renal excretion of bicarbonate secondarily increases, and the overall maternal blood pH remains relatively unchanged. I5 Except for the total protein and lactate levels, we found no significant variation in the coelomic fluid biochemical composition between 7 and 11 weeks of gestation and no correlation for the different variables between coelomic fluid and maternal serum, suggesting that the coelomic fluid acid-base regulation is not directly influenced by maternal physiologic changes. During that period the coelomic fluid had a lower pH, base excess, and bicarbonate level than did maternal venous blood and contained higher levels of Pc0 2 , lactate, and phosphate and lower levels of protein than did the maternal serum (Table I). These findings are consistent with a metabolic anaerobic acidosis and suggest the accumulation of acid bioproducts from the placental metabolism in the exocoelomic cavity.
Specific placental proteins such as human chorionic gonadotropin are found in higher concentrations in the coelomic fluid than in maternal serum, suggesting the presence of a preferential transfer of placental bioproducts toward the exocoelomic cavity. Jfi The significant drop in glucose level from maternal serum to the coelomic fluid probably reflects the important consumption of this substance by the placental tissue (Table I). More than half of the carbon dioxide produced by the gestational sac metabolism is removed in this form through the placenta. I7 Carbonic anhydrase activity has been demonstrated in the human villous tissue as early as the sixth week of gestation. 1" Experiments with animals have shown that there are at least two mechanisms for placental bicarbonate transfer (i.e., by conversion to carbon dioxide, which then diffuses across the villous barrier or by using an anion exchange system involving chloride and lactate ions I7 ). The latter mechanism could explain the high lactate concentration found in the coelomic fluid, which is known to be a major substrate for the energy required by the growing embryo or fetus. 2 High concentrations of phosphate in the coelomic fluid (Table I) suggest that this molecule serves as a buffer for the coelomic fluid acid-base balance. By contrast, the role of proteins is probably limited, because their level in the coelomic fluid were about 18 times lower than those of matched samples of maternal serum. These data suggest that the coelomic
Jauniaux et al.
Volume 170, Number 5, Part I Am J Obstet Gynecol
fluid composition is mainly influenced by the biologic activities of the placenta, which forms a continuous barrier to the maternal circulation and is the dominant embryonic organ at this time of gestation. During the second and third trimester the pool of amniotic fluid is subject to a constant turnover, being added to by fetal lung fluid and urine and being removed by fetal swallowing.' When the definitive placenta has formed, movements of water and electrolyte also take place across the free placental membranes. 3 The amniotic fluid collected before 11 weeks of gestation had a higher pH and base excess, higher levels of lactate and bicarbonate, and lower levels of total protein, phosphate, and chloride than did the coelomic fluid collected during the same period of gestation (Table I). Among these biologic parameters only pH and bicarbonate levels varied between 7 and 11 weeks of gestation. The metabolic alkalosis of the amniotic fluid observed in the first trimester probably arises from the accumulation of bicarbonate and the increased metabolism of organic anions such as lactate in the embryonic tissue. The embryonic skin, which only becomes keratinized during the second trimester,' is probably the major source of amniotic fluid in early pregnancy. The development of the ruinary and respiratory tracts and the rupture of the cloacal membrane during the eighth week of gestation must also influence the amniotic fluid composition from the third month of pregnancy. In particular, the mesonephros or temporary kidney that develops between embryonic stages 10 and 14 is capable of producing urine from embryonic state 17, which corresponds to 7.5 weeks after the last menstruation. 19 The metanephros or permanent kidney starts to form at embryonic stage 14 and produces urine around 10 weeks of gestation. 19 Significant changes were observed in the amniotic fluid composition at the end of the first trimester characterized by a decrease of pH (Fig. 1), base excess, and bicarbonate level and an increase of Pco 2 and chloride levels (Tables II and III). These changes of the amniotic composition continue during the second and third trimesters 8 - lo • 19 and probably reflect the increasing contribution of the fetal metaboli'sm to the amniotic fluid composition. It is of interest that except for the total protein level, which remained lower, the mean value of the other amniotic biologic parameters obtained after 11 weeks of gestation became similar to those found in the coelomic fluid before 11 weeks (Table I). We suggest that the changes of the amniotic fluid composition observed at the end of the first trimester correspond to the switch from the mesonephros to the metanephros and that from this stage of gestation acid bioproducts produced in increasing amounts by the metabolism of the rapidly growing fetus are mainly eliminated in the amniotic cavity by the immature kidneys.
1369
We thank the King's Health Care Authority and the staff of the Day Surgery Centre and Labour ward for their constant support. REFERENCES I. Sibley CP, Boyd RDH. Mechanism of transfer across the human placenta. In: Polin RA, Fox WW. eds. Fetal and neonatal physiology. Philadelphia: WB Saunders, 1992:6274. 2. Burd LJ, Jones MD, Simmonds MA, Makowski EL, Meschia G, Battaglia FC. Placental production and foetal utilization of lactate and pyruvate. Kature 1975;254: 710-2. 3. MacCarthy T, Saunders P. The origin and Circulation of the amniotic fluid. In: Fairweather DVI. Eskes TKAB, eds. Amniotic fluid: research and clinical application. 2nd ed. Amsterdam: Excerpta Medica, 1978:1-18. 4. BoydJD, Hamilton WJ. The human placenta. Cambridge: Heffer, 1970. 5. Jauniaux E, Burton GJ, Jones CP]. Early human placental morphology. In: Barnea E, Hustin J, Jauniaux E, eds. The first twelve weeks of gestation. Heidelberg: Springer-Verlag, 1992:45-64. 6. Jauniaux E, Jurkovic D, Gulbis B, Gervy C, Ooms HA, Campbell S. Biochemical composition of exocoeiomlC fluid in early human pregnancy. Obstet GynecoI1991;78: 1124-8. 7. Lind T. The biochemistry of amniotic fluid. In: Fairweather DVI, Eskes TKAB. Amniotic fluid: research and clinical application. 2nd ed. Amsterdam: Excerpta Medica, 1978:59-80. 8. Economides DI, Johnson MA, MacKenzie IZ. Does amniotic fluid analysis reflect acid-base balance in fetal blood. AM J OBSTET GYNECOL 1992;166:970-3. 9. Johnell HE, Nilsson BA. Oxygen tension, acid-base status and electrolytes in human amniotic fluid. Acta Obstet Gynecol Scand 1971 ;50: 183-92. 10. Sinha R, Carlton M. The volume and composition of amnotic fluid in early pregnancy. J Obstet Gynaecol Br Commonw 1970;77:211-4. II. Jurkovic D, Jauniaux E, Campbell S, Pandya P, Cardy DL, Nicolaides KH. Coelocentesis: a new technique for early prenatal diagnosis. Lancet 1993;341: 1623-4. 12. HustinJ, SchaapsJP. Echographic and anatomic studies of the maternotrophoblastic border during the first trimester of pregnancy. AM J OBSH:T GYNECOL 1987; 157: I 62-8. 13. Hustin.J, Schaaps .JP, Lambotte R. Anatomical studies of the uteroplacental vascularization in the first trimester of pregnancy. Trophoblast Res 1988;3:49-60. 14. Rodesch F, Simon P, Donner C, Jauniaux E. Oxygen measurements in the maternotrophoblastic border during early pregnancy. Obstet Gynecol 1992;80:283-285. 15. Lim VS, Katz AI, Lindheimer MD. ACId-base regulation in pregnancy. AmJ PhysioI1976;231:1764-70. 16. Jauniaux E, Gulbis B, Jurkovic D, Schaaps JP, Campbell S, Meuris S. Protein and steroid level, in embryonic cavities of early human pregnancy. Hum Reprod 1993;8:782-7. 17. Hatano H, Leichtweij3 HP, Schroder H. Uptake of bicarbonate/C0 2 in the isolated guinea pig placenta. Placenta 1989; 10:212-21. 18. Aliakbar S, Brown PR, Jauniaux E, Bidwell DE, Kicolaides KH. Measurements of carbonic anhydrase Isoenzyme, in early human placental tissues. Biochem Soc Tram 1990; 18:670. 19. O'Rahilly R, Muller F. Human embryology and teratology. New York: Whiley-Liss, 1992.