Blood ammonia concentration in cord blood during pregnancy

Early Human Development, 33 (1993) l-8

@ 1993 Elsevier Scientific Publishers Ireland Ltd. All rights reserved. 0378-3782/93/$06.00 EHD 01385

Blood ammonia concentration in cord blood during pregnancy Janice T. DeSanto, Wallace Nagomi, Edward A. Liechty and James A. Lemons Department of Pediatrics, Section of Perinatal-Neonatal Medicine, Indiana University School of Medicine and the James Whitcomb Riley Hospital for Children, Indianapolis, Indiana (USA)

(Received 25 June 1992; revision received 28 December 1992; accepted 5 January 1993)

In vivo studies in several animal species and in vitro studies of human placenta suggest that ammonia is produced within the uteroplacenta and released into the fetal and maternal circulations. Furthermore, the fetal blood ammonia levels in sheep have been found to be signiticantly greater than maternal levels and inversely correlated with gestational age. Our present study had two main goals: first, to assess whether ammonia may be produced in the human placenta and taken up by the fetus and second, to determine if the fetal blood ammonia levels are greater than maternal and inversely correlated to gestational age. We measured the blood concentrations of ammonia by obtaining both umbilical venous (UV) and arterial (UA) samples obtained from doubly clamped sections of umbilical cord at the time of delivery. Blood gases and glucose were also measured on each sample. Samples were obtained at 96 deliveries (70 vaginal, 26 cesarean section, 18/26 cesarean deliveries not in labor). The deliveries were separated into two groups. Group A included all deliveries, both vaginal and cesarean section, while Group B consisted of only cesarean section prior to onset of labor. The mean gestational age for all deliveries was 38.4 weeks f 2.95 (S.D.) with a range of 25-43 weeks of gestation. There was no significant relationship between the fetal ammonia levels and gestational age in either group. In both groups A and B, the UA ammonia concentrations were significantly higher than the paired UV ammonia levels (UV-UA: Group A: -10.00 + 17.6, Group B: -9.3 f 12.6). Conclusion: These data show no correlation between fetal ammonia levels and gestational age. However, umbilical ammonia concentrations are approximately twice normal adult levels. A negative UV-UA ammonia difference in the Correspondence to: James A. Lemons, Department of Pediatrics, Indiana University School of Medicine, 702 Barnhill Drive, R208, Indianapolis, IN 46202, USA.

2

fetal circulation disputes uteroplacental production of ammonia, although this Iinding may reflect the stress of delivery and not steady state in utero conditions. Alternatively, the human fetus may behave differently from the sheep fetus. Key words: ammonia production; fetal metabolism; uteroplacental umbilical artery; umbilical vein; fetus; human

metabolism;

Introduction Ammonia, traditionally thought to be toxic to developing tissue, has been found to be important in normal fetal metabolism. In vitro and in vivo animal studies have shown that ammonia is produced within the uteroplacenta and released into the fetal In addition, fetal and maternal circulations simultaneously [2,6,8- 10,l 5,16,19,20]. blood ammonia concentrations in sheep and other species are significantly elevated as compared to maternal levels and the concentrations have been found to be inversely correlated to gestational age [2,8]. In contrast to data obtained in chronic animal preparations, clinical studies in humans have found a higher umbilical arterial ammonia concentration than that in the umbilical vein, suggesting net excretion of ammonia out of the fetus into the uteroplacenta [20]. This discrepancy may be a reflection of the fact that the clinical studies are necessarily carried out in acute nonsteady state conditions at the time of parturition which may not reflect the normal in utero environment. Alternatively, it is possible that the developmental process differs in the human fetus as compared to sheep or other species. Our study was designed to test several hypotheses. We hypothesized that ammonia is produced in the human placenta and transferred to the fetus in net quantity; and that the fetal blood ammonia levels are inversely correlated to gestational age in human pregnancy. In addition, we postulated that the stress of labor and vaginal delivery may cause alterations in the net balance of ammonia, while cesarean section prior to labor would more accurately reflect in utero conditions. Methods This investigation was approved by the Institutional Review Board of Indiana University - Purdue University of Indianapolis. Informed consent was not required as there were no known risks to the infant or mother associated with the study. Patients were randomly selected among those deliveries at the Indiana Medical Center. Deliveries were attended by one of the investigators (WN or JD) weekdays between 0800 and 1600 h. We separated the deliveries into two groups. Group A included all deliveries, both vaginal and cesarean section, while Group B consisted of only cesarean section deliveries to women not yet in labor. Group B was intended to reflect in utero values more closely.

3

Umbilical venous (UV) and arterial (UA) samples were obtained from doubly clamped sections of umbilical cord at the time of delivery of the infant. Deliveries were excluded if the cords had been clamped for more than 3 min or had already been utilized for routine cord blood gases. Approximately 3 cm3 of blood were placed into iced sodium heparinized tubes and 0.5 cm3 of blood into iced heparinized syringes. Analyses were then performed within 30 min to measure ammonia, glucose and blood gas. Plasma ammonia was measured on a DuPont Automatic Clinical Analyzer (ACA) TM11 and III (DuPont Comp, Wilmington, DE) by an enzymatic method. This method depends upon glutamate dehydrogenase catalyzing the condensation of ammonia and a-ketoglutarate with simultaneous oxidation of reduced NADPH. The decrease in absorbance at 340 nm secondary to disappearance of NADPH over a specified time period is directly proportional to the ammonia concentration in the sample. The intra-assay coefficient of variability in our laboratory is 2.8%. Whole blood glucose was measured using a glucose oxidase method (Ektachem Clinical Chemistry Slide) with an intra-assay coefficient of variability of 2.9%. The venous and arterial pH, PO,, Pco2 were measured using a BGM TM system 1312 ph/blood gas analyzer (Instrumental Laboratory System; Lexington, MA). Because of variations in the blood volume obtained, it was not possible to obtain a complete set of samples in every instance. Statistical Methods The results are expressed as mean f SD., frequency, and range. The paired r-test was used to calculate the significance of the differences between UV and UA concentrations. Regression analyses were carried out by the least squares method. Results Blood samples were obtained from 96 deliveries (Group A). Seventy of these deliveries were vaginal (73% of all deliveries) while 26 were cesarean section (27% of all deliveries). The mean gestational age for all deliveries was 38.4 weeks f 2.95 (S.D.) with a range of 25-43 weeks gestation. The length of labor ranged from 0 to 48 h. Abnormal fetal heart rate patterns consistent with fetal distress were noted in nine cases (four - late deceleration, five - variable deceleration) while meconium stained amniotic fluid was present in 11 deliveries. All infants except for 10 had I-min Apgar scores of 7 or greater. Three infants, all delivered by cesarean section prior to onset of labor, had significant malformations, including single cases of hydrocephalus, dextrocardia, and multicystic kidneys. Group B consisted of 18 women who underwent cesarean section prior to onset of labor. The mean gestational age for Group B was 38.0 weeks f 3.76(S.D.) with a range of 26-42 weeks gestation. Two pregnancies were complicated by preeclampsia and three pregnancies were complicated by either gestational diabetes or insulin dependent diabetes. Three infants had significant malformations as described above. All infants except for six had I-min Apgar scores of 7 or greater. The mean UA and UV ammonia and glucose concentrations and blood gases for

I

(mg/dl)

arterial

PC02

F-02

Group B NH, (/.W Glucose (mg/dl) PH

PC02

PO2

Glucose PH

.Group A NH3 W)

Mean umbilical

TABLE

S.D.

Mean

52;2 77.9 7.32 26.7 45.9

64.6

f 0.05 zt 8.9 f 5.3

?? 9.7

f

17 17 15 15 15

88

20.2 0.06 7.0 6.3

92 88 88 87

zt zt zk zt

89.2 7.35 26.6 42.1

n

49.3 zt 54.5

f

vein

Umbilical

and venous ammonia

25.0-301 60.0-94 7.18-7.36 10.0-44.0 27.0-59.0

47.0-164.0 7.14-7.49 10.0-45.0 28.0-64.2

o-452.0

Range

and glucose

63.7 67.7 7.27 16.5 54.5

78.2 7.29 18.8 50.9

f f f f *

f f zt f

63.8 f

Mean

f

Umbilical

concentrations,

78.4 9.9 0.06 5.2 7.5

18.9 0.07 7.1 9.5

71.9

S.D.

artery

15 13 12 12 12

67 55 55 55

66

n

30.0-345.0 49.0-87.0 7.14-7.34 10.0-28.0 41.5-69.2

49.0-136 7.13-7.53 5.0-38.0 17.0-73.0

15.0-516

Range

blood gases and umbilical

-9.33 11.92 0.64 12.5 -10.4

10.42 0.51 11.2 -8.5

-10.00

VA

zt 17.6

difference

differences

f zt f f f

12.57 5.39 0.04 6.5 6.0

zt 9.99 f 0.08 f 7.9 f 10.9

Umbilical

venoarterial

15 12 12 12 12

66 53 53 52

64

n

A and B.

-44.0-16.0 4.0-21.0 0.02-o. 16 5-30 -20.9-0.4

-14.0-49.0 -0.39-0.19 -11.0-33.5 -36.2-47.2

-64.0-49.0

Range

in Groups

CO.0122 CO.0122 CO.0001

co.001

P-value

5

both Groups A and B are reported in Table I. The range of UA ammonia concentration was 15-516 PM with a mean of 63.7 PM. Three cases in Group A resulted in UA ammonia levels exceeding 102 PM. The first case with an ammonia level of 193 PM was an uncomplicated delivery after 12 h of labor. The second case with a level of 345 PM was an uncomplicated elective C-section for a post term infant. The last case with the highest level of 5 16 PM was significant for chorioamnionitis along with prolonged labor of 48 h ending in emergency C-section for fetal distress. Resuscitation and intubation were required and the infant’s Apgar score were 2 at 1 min and 5 at 5 min. The data were analyzed to test for a relationship between gestational age (GA) and the ammonia concentration in the UA and UV samples. No relationship could be demonstrated in either Group A or B, even if the three highest values (> 100 PM) were excluded from analyses (Group A UA vs. GA, r2 = 0.0001, UV vs. GA, r* = 0.013; Group B UA vs. GA, r2 = 0.005, UV vs. GA, r2 = 0.024). Secondly, no relationship between gestational age and umbilical venoarterial (UV-UA) concentration differences was seen for either group A or B (Group A, r2 = 0.0001; Group B, r2 = 0.003). In Group A, the UA ammonia concentration was significantly higher than the paired UV ammonia concentration (n = 64, UA 62 PM, UV 52 PM, P < 0.001). In Group B, again the UA ammonia concentration was significantly higher than the paired UV ammonia concentration (n = 15, UA 64 PM, UV 54 FM, P c 0.0122). UV ammonia concentration was higher than UA concentration in 12 of 64 deliveries in group A and in one of 15 deliveries in group B. Both groups demonstrated a negative mean UV-UA ammonia difference (Group A -10.0 f 17.6, Group B -9.3 f 12.6). The UA glucose concentrations in both groups were significantly lower than the paired UV glucose concentrations (Group A, n = 66, 78.5 vs. 88.9 mg/dl (UA vs. UV), P < 0.0001; Group B, n = 17, 67.7 vs. 77.9 mg/dl (UA vs. UV), P < 0.0122). Both groups have positive mean UV-UA glucose differences (Group A, + 10.4 mg/dl; Group B, +11.9 mg/dl). The UV pII and PO, values were significantly higher than the paired UA values in both groups. As expected, the UV Pc02values were consistently lower than the paired UA values in both groups. Discussion

The fetus acquires nutrients from the maternal plasma while excreting metabolic byproducts to the maternal compartment via placental transfer. It has become increasingly important to define the relationships between the placenta, the fetus, and the mother in the exchange of these various substrates. Valuable information concerning mammalian fetal metabolism has been derived from extensive investigations of the chronic fetal sheep preparation. Results obtained in the ovine fetus indicating uteroplacental production of ammonia with resultant high fetal concentrations prompted the present investigation. Our present study had two main goals: first, to assess whether ammonia is produced in the human placenta and taken up by the

6

fetus; and second, to determine if the fetal blood ammonia levels are inversely correlated to gestational age. Previous studies of the unstressed ovine fetus indicate that ammonia is produced in significant quantities throughout pregnancy by the uteroplacenta and delivered into both the umbilical and uterine circulations [8,9]. Because of the consistent infusion of ammonia into the umbilical circulation, fetal ammonia levels are significantly greater than maternal and inversely related to gestational age [2,8]. In vitro studies of human placenta also demonstrate ammonia production [ 151. In vivo studies of guinea pig and rabbit also document ammonia excretion from the uteroplacenta into the maternal circulation [ 1, lo]. Results of the present investigation differ from observations made in the ovine fetus. Our study failed to demonstrate positive UV-UA differences as noted in the chronic catheterized sheep models (Group A, -10.0; Group B, -9.3). Our results are similar to those reported by Rubaltelli, also obtained in human subjects from umbilical cord samples at the time of delivery. The observations of Rubaltelli demonstrated a negative UV-UA difference of -3.49 PM (UV 29.16 PM, UA 32.62 PM) [20]. However, we were able to reproduce expected values regarding glucose and blood gases in the UV and UA samples [3,21]. Although variability between samples was large, these findings suggest that sample collection was done under reasonable conditions. It is possible that length of labor, method of delivery, type of anesthesia, and/or other variables will alter fetal ammonia levels from basal fetal concentrations. It is difficult to obtain samples from human pregnant subjects during steady state in utero conditions, but may be possible by either fetoscopy or percutaneous umbilical blood sampling. We interpret our present findings and those of previous clinical studies to reflect most likely the physiologic stress of labor and delivery on the fetus. Alternatively, the human fetus may behave differently from the sheep fetus and may produce ammonia and excrete it into the placenta. This may be a normal developmental process in the human fetus. We were also unable to document an inverse relationship between fetal ammonia concentrations and gestational age in both groups in our present study. This again may reflect the acute stresses surrounding delivery as well as the relatively small number of very premature infant samples. Nonetheless, the absolute concentrations of ammonia in the umbilical circulation are significantly higher than normal pregnant and nonpregnant adult values. The levels are appoximately twice those observed in adulthood [20]. It should be noted that after birth, the concentrations of ammonia decrease with age. The ammonia concentrations in the newborn and children are higher than those of the adult. Thus these findings may be due to normal maturation of ammonia metabolism and excretion. Several important questions remain unanswered with regard to ammonia metabolism within the conceptus. The role of ammonia in maintaining fetal and placental nitrogen balance is unclear. Although ammonia may represent a significant avenue for nitrogen excretion from the uteroplacenta, there is no evidence that the capacity for urea production, the major form of excess nitrogen excreted, by the fetus is ever exceeded. It is possible that the persistent ammonia excretion from the uteroplacenta in later gestation may be a carry over from earlier in gestation when urea synthesis

was not a functional process yet [ 181. Ammonia may also represent a method of hydrogen ion secretion throughout gestation to maintain acid-base balance [22]. The origin of the ammonia produced in the conceptus is also not known. While amino acids represent a likely source in view of the rich supply of amino acid transaminases and deaminases in the placenta, studies have not been performed which trace nitrogen metabolism within the placental compartment [5,12]. Generation of substantial quantities of ammonia is also possible during pyrimidine metabolism. Degradation of adenosine triphosphate occurs at a rapid rate in the placenta which is perhaps the most metabolically active organ of all human tissues. It should be noted that ATP metabolism is the chief source of ammonia for another metabolically active tissue, exercising skeletal muscle [ 171. Perhaps the most important question concerning ammonia is the potential effect on fetal and placental metabolism throughout pregnancy when present at such high concentrations. Traditionally, ammonia when present in high concentrations has been considered both a cytotoxin and neurotoxin. Increased concentrations of ammonia have also been thought to reflect aberrations in metabolism and loss of integrity of normal cell function [4,7]. Therefore, it is surprising to find that fetal levels in animals and in humans are significantly higher than maternal concentrations throughout pregnancy. It is likely that such modest elevations in ammonia concentration will influence fetal metabolism, as is true for adult liver. Nonetheless, the exact role of ammonia in fetal and placental metabolic regulation remains unknown. References 1 2 3 4 5

6

Battaglia, F.C. and Meschia, G. (1986): An Introduction to Fetal Physiology, p. 114. Academic Press, Orlando, Florida. Bell, A.W., Kennaugh, J.M. et al. (1989): Uptake of amino acids and ammonia at midgestation by the fetal lamb. Q. J. Exp. Physiol., 74, 635-643. Comline, R.S., Silver, M. (1976): Some aspects of foetal and uteroplacental metabolism in cows with indwelling umbilical and uterine vascular catheters. J. Physiol., 260, 571-586. Flannery, D.B., Hsia, Y.E. and Wolf, B. (1982): Current status of hyperammonemic syndromes. Hepatology, 2, 495-506. Goodwin, G.W., Gibboney, W., Paxton, R. et al. (1987): Activities of branched chain amino acid aminotransferase and branched chain 2-oxoacid dehydrogenase complex in tissues of maternal and fetal sheep. B&hem. J., 242, 305-308. Hauguel, S., Challier, J.C. et al. (1983): Metabolism of the human placenta perfused in vitro: glucose transfer and utilization, O2 consumption, lactate and ammonia production. Pediatr. Res., 17, 729-732. Hindfelt, B., Plum, F. and Duffy, T.E. (1977): Effect of acute ammonia intoxication on cerebral metabolism in rats with portacaval shunts. J. Clin. Invest. 59, 386-396. Holxrnan, I.R., Lemons, J.A:, Meschia, G. et al. (1977): Ammonia production by the pregnant uterus. Proc. Sot. Exp. Biol. Med., 156, 27-30. Holzman, I.R., Phillips, A.F. and Battaglia, F.C. (1979): Glucose metaolism, lactate, ammonia production by the human placenta in vitro. Pediatr. Res., 13, 117-120. Johnson, R.L., Gilbert, M. et al. (1986): Uterine metabolism of the pregnant rabbit under chronic steady-state conditions. Am. J. Obstet. Gynecol., 154, 1146-51. Koch, G. and Wendel, H. (1968): Adjustment of arterial blood gases and acid base balance in the normal newborn infant during the first week of life. Biol. Neonate, 12, 136-161. Liechty, E.A., Barone, S. and Nutt, M. (1987): Effect of maternal fasting on ovine fetal and maternal branched chain amino acid transminase activities. Biol. Neonate, 52, 166-173.

8 13 Low, J.A. (1966): Assessment of metabolic state and anaerobic metabolism in the normal newborn infant. Am. J. Obstet. Gynecol., 94, 497-505. 14 Low, J.A., Pancham, S.R., Worthington, D. and Boston, R.W. (1974): Acid-base, lactate, and pyruvate characteristics of the normal obstetric patient and fetus during the intrapartum period. Am. J. Obstet. Gynecol. 120, 862-867. Luschinsky, H.L. (1951): The activity of glutaminase in the human placenta. Arch. Biochem. Biophys., 31, 132-140. Meschia, G., Battaglia, F.C. et al. (1980): Utilization of substrates by the ovine placenta in vivo. Fed. Proc., 39, 245-249. Meyer, R.A. and Terjung, R.L. (1979): Differences in ammonia and adenylate metabolism in contracting fast and slow muscle. Am. J. Physiol., 237, Clll-C118. Raiha, NCR. and Kekomaki, M. (1975): Developmental aspects of ammo acid metabolism in the human. In: Total Parenteral Nutrition, pp. 208. Editor: H. Ghadimi, John Wiley and Sons, Inc., New York. 19 Remesar, K., Arola, L., Palou, A. et al. (1980): Activities of enzymes involved in amino acid metabolism in developing rat placenta. Eur. J. B&hem., 110, 289-293. 20 Rubaltelli, F.F. and Formentin, P.A. (1968): Ammonia nitrogen, urea, and uric acid blood levels in the mother and in both umbilical vessels at delivery. Biol. Neonate, 13, 147-154. 21 Stembera, Z.K. and Hodr, J. (1966): I. The relationship between the blood levels of glucose, lactic acid and pyruvic acid in the mother and in both umbilical vessels of the healthy fetus. Biol. Neonate, 10, 227-238. 22 Tannen, R.L. (1978): Ammonia metabolism. Am. J. Physiol.: Renal Fluid Electrolyte Physiol., 4, F265-F277.