Comparative analysis of normal and growth-retarded placentas with phosphorus nuclear magnetic resonance spectroscopy Helen H. Kay, MD,"< Saralyn R. Hawkins, MD: John D. Gordon, MD: Yuping Wang, MD: Anthony A. Ribeiro, PhD,< and Leonard D. Spicer, PhDb .< Durham, North Carolina OBJECTIVE: Phosphorus 31 magnetic resonance spectroscopy studies were carried out on placentas from normal vaginal and elective cesarean deliveries without antenatal complications and from pregnancies complicated by intrauterine growth retardation of unknown cause to determine differences. STUDY DESIGN: Perchloric acid extraction was performed on frozen tissue. and quantitative analysis was carried out for well-resolved resonances representing adenosine triphosphate, sugar phosphate, inorganic phosphorus, diphosphoglycerate, glycerophosphorylethanolamine, and glycerophosphorylcholine. RESULTS: Adenosine triphosphate levels were highest in the growth-retarded group. There were significantly higher levels of sugar phosphate, diphosphoglycerate, and glycerophosphorylcholine in the placentas of the growth-retarded pregnancies compared with those from normal placentas. CONCLUSION: These differences may represent a response to hypoxia and an increase in the amount of blood in the placenta. The results demonstrate the utility of nuclear magnetic resonance spectroscopy for studying the pathology of abnormal placentas to gain a better understanding of the pathology and represent early steps toward in vivo spectroscopic studies of the placenta . (AM J OSSTET GVNECOL 1992;167:548-53.)
Key words: Phosphorus 31 nuclear magnetic resonance spectroscopy, human placenta, perchloric acid extraction, intrauterine growth retardation Intrauterine growth retardation (IUGR) is a complication of pregnancy affecting as many as 4% to 7% of all pregnancies.' The causes of fetal growth retardation include maternal hypertension, cigarette and other substance abuse, in ut ero infection, and chromosomal disorders. In many cases, however, a specific cause is never identified. The placenta, which serves many critical functions for the fetus, is an important organ to study in any attempt to understand the cause of idiopathic IUGR. Very little is known about the role of the placenta in fetal IUGR. Most studies have focused on pathologic and histologic examinations. These investigations, however, tended to be nonspecific because of the many different causes of IUGR.2 Other studies on animals have reported a decrease in glucose and amino acid transfer to the fetus. "" Few stud ies, however, ha ve yet focused on a biochemical analysisof placental metabolic fun ction. Such analysis may be particularly informative in cases where the cause is unidentified.
From the Departments of Obstetrics and Gynecology: Biochemistry,' and R adiology: Duke University M edical Center. Supp orted by National Institutes of Health grant N o. 1K08HD00822-02 (H. K. ). R eceioedfor publication August 4, 1991; revisedJanuary 15,1992; accepted March 31,1992. R eprint requests: Helen H. Kay, MD, Box 3373, Duke University M edical Center, Durham , NC 27710. 611138320
548
Nuclear magnetic resonance (NMR ) spectroscopy offer s a unique method for studying placental metabolic fun ction. It is a noninvasive technology that simultaneously monitors naturally occurring molecules in biologic tissue . In a previous report" we described the phosphorus 31 magnetic reson ance spectra of fresh human placental tissues and perchloric acid-extracted tissue. We also detailed the quantitative analysis of adenosine triphosphate and inorganic phosphorus levels in the placental extracts. In the current study we extend our quantitative analy sis to other metabolites easily visible in the phosphorus spectra, such as sugar phosphate, diphosphoglycerate, glycerophosphorylethanolamine, and glycerophosphorylcholine. We have focused our anal ysis on a compa rison of placentas from idiopathic growth-retarded pregnancies and normal pregnancies to gain further insights into the cause of the growth retardation.
Material and methods Five placentas from pregnancies antenatally identified as complicated by IUGR of unknown cause were obtained at the time of delivery by cesarean section. T hese were confirmed as growth-retarded placentas by neonatal weights < 10th percentile with Brenner's growth curve.' The mean birth weight was 1253 ± 207 gm at a mean gestational age of 35 .3 ± 0.9 weeks con firmed by ultrasonographic examination before 20 weeks . Thorough chart review and patient interview
NMR spectroscopy of normal and growth-retarded placentas
Volume 167 Nu mb er 2
revealed no specific cause suc h as hypertension, cigarette or drug abuse, viral illnesses, or prenatal complications , i.e., bleeding. The mean placental weight was 198 :t 32 gm, and there were no gross abnormalities such as infarctions or abruptio placentae seen on inspection. Small pieces of placental tissue were removed from multiple regions on th e mat ernal side by sharp dissection and placed into liquid nitrogen within I min ute a nd sto red in a-70° C free zer for later anal ysis. Similarl y, placentas from normal, un complicated pregnan cies obtained from eight vaginal deli veries and eight elective cesarean section deliver ies were harvested and fr ozen in the same fashion . Placentas in all three groups were obtained in the third tr imester. At the time of extraction the tissue was removed from the freezer and pulverized in liquid nitrogen with a stainless steel Waring blender (Dynamics Corp. of America, New Hartford, Conn .). A perchloric acid extract was performed as pr eviou sly des cribed." Briefly, a 20 gm sample of finely gro und placental tissue was mixed with 20 ml of I mol / L perchloric acid that had been precooled to -78 C in dry ice. The perchloric acid - tissue mixture was allowed to warm to approximatel y - 10° C in the ambi en t air for 15 minutes during which time the metabolites wer e extr acted . The extract was then centrifuged at 40 ,000g at _4° C for 20 minute s to separate the solid fr action from the liquid fra ction . T he liquid supernatant conta ining the extracted metabolit es was then decanted into a chilled beaker a nd neutralized to pH 7.0 at 4° C through the dropwise add ition of potass ium hydroxide. The potassium perchlorate precipitated durin g neutralization was re moved by a second centrifugation step identical to the first, and the supernatant was then passed over a Chelex- l 00 chelating resin (Sigma Chemical Co., St Loui s) to remove divalent cations. The extract was again titr at ed to neutral pH with diluted perchloric acid, frozen in a vacuu m flask (Virtis Co ., Gardiner, N.Y.), and lyophilized to dryness. The extract was then resuspended in 5 ml of 40 % deuterated water for quantitat ion of spectra. The extracts were placed in the outer chamber of a coaxial 10 mm NMR tube (Wilmad Glass, Buen a, N.J.) with an external reference standard of 25 mmol / L hexachlorocyclotriphosphazene mixed with 3 1 mmol /L chro mium acet ylaceto nate (Ald rich Chemical lnc., Milwaukee)" in an inn er coaxial insert. Spectra were obtained with a 21 usee pul se giving a 70° flip angle, an acquisition period of 1.12 seconds, a relaxation dela y of 15 seconds (total pulse recycle time 16.12 seconds), and 160 acqui sition s. When phosphorus nuclei have su fficient time to return to their equilibrium state , i.e., full y relaxed state , their peak intensities are proportional to the concentration of the metabolite present. To assess the consequence of the pulse recycle time (acquisition time of 0
549
1.12 seconds plus vari able del ay) o n the extent of nuclear relaxation, spectra were obtained at rec ycle times of6 .12, 11.12 , 16.12 ,21.12, 31.1 2, and 61.12 seconds. T he hexachlorocyclotriphosphazene and (3-adenosine triphosph ate resonances are fully relaxed under con dition s of our "sta ndard" 16.12-second recycle time." The suga r phosphate, ino rgani c phosphate, glyceroph osphorylethanolamine, and glycerophosphorylcholine resonances, respectively, return to 93.7%, 97 .4%, 95 .1%, and 98 .6% of their equ ilibrium states at the 16.12-second recycle time . The following "satura tion correction factors " are then used to correct the integra ted intensities of the signals obtained at 16.12 seconds: adenosine triphosphate 1.00, hexachlorocyclotriphosphazene 1.00; sugar ph osphate 1.07, inorganic ph osphate 1.03; glycerophosphorylcholine 1.05, and glycero phosphorylethano lamine 1.0I before calculation of metabolite concentrations . T he quantitative anal ysis is previou sly reported." Metabolite quantitation was performed with the external re feren ce standard of hex achlo rocyclotriphosphazene and chro miu m acet ylaceton ate as a spin relaxation agent (Ald rich) in a coaxial capillary system. In br ief, the integ rated peak are a of a ph osphorus resonance was divided by the integrated peak area of the hexachloroc yc1otriphosphazene reference signal. The rati o for th e (3-adenosine triphosphate re sonance is then used to qu antitate the adenosine triphosphate metabolite present with a calibration cu rve previously constr ucted for varying known concen tra tions of adenosine triphosphate. The ad en osine tr iphosphate levels and the (3-adenosine triphosph ate /hexachlorocyclotriph osphazene ratio are related by the equation }' = 0.035 + O.088X. After calculation of the concentr ation of adenosine triphosphate, the remaining peak ratios and their corresponding satu ration correction factor s were then used to determine the final concentr ation of the other metabolites. All spe ctr a were obtained on a 300 MHz NMR spectrometer (model GN-300 WB General Electric , Fre mont, Calif.) interfaced with a wide-bore superconducting magnet (89 mrn bore, 7.0 tesla field) in the Duke Magnetic Resonance Spectroscopy Center. A standa rd 20 mm broad band probe (General Electric) tu ned to 121.4 MHz was used for SI p spectroscopy. Spectroscopic data were collected at the ambient op erating probe temperature (approximately 22° C) without temperature regulation. T o avoid radiofrequency heating effects on the tissue a nd extracts, no proton decoupling was used . :lI p chemical shifts are referenced to phosphocreatine at 0.00 parts /m illion. Stat istical analysis was accompli shed with a Statgraphics program (Statistical Graphics Corporation, STSC, Rockville, Md.). The means and standard deviation of each metabolite were calculated for the three
550 Kayet al.
August 1992
Am
J Obstet Cynecol
Pi
HCCTP
SP
GPC
DPG Y-ATP
a-ATP
B-ATP
I -20
I'
10
-30
PPM
Fig. 1. Phosphorus NMR spectrum of perchloric acid-extracted placental tissue from normal vaginal delivery. Note resonances for external reference standard hexachlorocyclotriphosphazene (HCCTP), diphosphoglycerate (DPG), sugar phosphate (SP), inorganic phosphate (Pi), glycerophosphorylethanolamine (GPE), glycerophosphorylcholine (GPC), phosphocreatine (PCr), and adenosine triphosphate (A TP). groups (vaginal delivery, cesarean section, and growth retardation). A one-way analysis of variance was performed to determine the presence of a statistical difference among the means of the three groups. A multiple-range analysis or a t test was performed to determine specific differences among each of the three groups. Results
Fig. I shows the phosphorus spectrum from a perchloric acid extract of a normal placenta. Of note are the distinct and well resolved resonances for hexachlorocyclotriphosphazene, sugar phosphate, diphosphoglycerate, inorganic phosphorus, glycerophosphorylethanolamine, glycerophosphorylcholine, and the three phosphorus nuclei of adenosine triphosphate. Note that there is very little phosphocreatine at 0 parts/million, and there is a large amount of inorganic phosphorus. Table I lists the placental metabolite concentrations determined by 3lp NMR for vaginal-delivery placentas, cesarean-section placentas and growth-retarded placentas delivered by cesarean section. The table specifically lists the mean concentration and standard deviation for each of the metabolites of the three placenta
groups. The mean concentrations of the phosphorus metabolites are highest in the IUGR group, by almost a factor of two compared with the vaginal and cesarean section groups, except for inorganic phosphorus. There are no statistically significant differences among the metabolites between the vaginal and cesarean section placentas. However, there are statistically significant differences among the adenosine triphosphate, sugar phosphate, diphosphoglycerate, and glycerophosphorylcholine levels between the growth-retarded placentas and those from normal vaginal and cesarean section deliveries. Comment
The current study is a comparative analysis of placentas from normal and abnormal pregnancies with phosphorus NMR spectroscopy. Our findings suggest that NMR spectroscopy, with its ability to simultaneously display the levels of several metabolites, can be extremely useful for in vitro and in vivo biochemical analysis. In this study the NMR analysis yields further insights into placental metabolism. Human in vivo NMR studies in diagnostic medicine are currently focused on ischemic brain injury, muscle dystrophy, and tumor response to chemotherapy." With continued de-
Volum e 16i Number 2
NMR spectroscopy of normal and growth-retarded placentas 551
Table I. Placental metabolite concentrations determined by phosphorus NMR spe ctroscopy
Metabolit e
Adenosine triphosphate Sugar phosphate Inorgan ic phosphate Diph osphoglycerate Glycerophosphoryleth an olamine Glycerophosphorylcholin e
No rmal vaginal delivery (n = 8) (mean ± S D) (um ol! gm wet weight)
0.270 1.l03 2.586 0.383 0.273 0.556
± 0.189 ± 0.308 ± 0.308
± 0.135 ± 0.095 ± 0 .139
No rmal cesarean section (n = 8) (mean ± S D) (umol ]gm wet weight)
0.426 1.502 3.477 0.4 26 0.347 0.685
± 0.154
± 0.305+
± 1.643
± 0.151+
± 0.084
± 0. 130:j:
Growth retardation (n = 5) (mean ± SD) (u mol! gm wet weight)
0.547 2.137 3.720 0.712 0.440 1.112
± 0.160t
0.31 9§ 1.777 0.090§ 0.16 6 ± 0.348§
± ± ± ±
p Valu e*
0.0295 0.0001 0.2740 0.001 0 0.0528 0.0005
*T he p values obtained fro m a one-way ana lysis of vari an ce to test differ en ces between means in all three gro ups . t Statistically significant difference between the growth-retarded and normal vaginal d eliver y groups by a t test but not th e multiple range test. :j:Statistically significant di fference betwee n the growth-retarded and normal cesarean section groups by mul tiple range test. §Statistically significant difference between the growth-retarded and normal vaginal deliv er y gro ups by multiple range test.
velopments in technology, detailed an alysis ma y someda y be feasible for in vivo spectrosco pic studies of the human placenta. Our current report focu ses on differences bet wee n spectra fr om tissue extracted with perchloric acid both for th e increased resolution available to identify biochemical change and with the expectation that the se differences, once established, may be visible in whole placental tissue. It is an important first step to characterize th e NMR spectra of human placental extracts to facilitate the analy sis of peaks in the spectra from gross tissue for relevant biochemical and diagnostic studies in the future . Our previous preliminary studies on gro ss placental tissue" demonstrated obvious resonances for sugar phosph ate, inorganic phosphorus and adenosine triphosphate. The diphosphoglycerate, glycerophosphorylethanolam ine , and glycerophosphorykholine resonances were not well resolved. In addition, phosphocreatine was absent in the spectrum of gross tissue. However, with perchloric acid extraction we were able to con firm the presence of phosphocreatine albeit in very small qu antities. We were also able to determine tissue ad en osine triphosph ate levels from the extracts. In this stu dy we performed further analyses on the spectra from perchlori c acid-extracted tissue . We found different ad enosine triphosphate levels in the three groups studied. They were lowest in the vaginal delivery group, although not statistically different from the cesarean section gr oup. The adenosine triphosphate level is highest in the IUGR group with a statistical difference from the vaginal delivery group. Bloxam 10 also reported lower levels of adenosine triphosphat e in placentas from vaginal deliveries than from placentas from cesarean section. This may, in part, be due to varying a mounts of blood within the placentas. Less blood is expected in vaginal delivery placentas because blood is squeezed out of the sinuses during labor and delivery. We have previously found"
that the blood concentration within th e placentas is significant and may accou nt for as mu ch as 30% to 50% of the total amou nt of ade nosine triphosphate seen. Therefore the fact that the highest level was found in the I UGR placentas is interesting and suggests that these placentas may have th e largest amount of blood . This is further supported by the fact th at diphosphoglycera te , which is a major component of blood, is present at th e highest level in the IUGR gr oup and is statistically di fferent from the other two groups. This increase in blood content in the IUGR placentas may represent a com pensatory increase in blood flow. It is also possible th at differences in the adenosine triphosphate levels re sult from varying oxidative phosphorylation requirements within the placentas at the time of delivery. Levels are likely to be lower in the vaginal delivery group because the adenosine triphosphate is used during the process of labor. T he increased levels in placentas from growth-retarded pregnancies, on the other hand, could arise from a need to increase oxidative phosphorylation to maintain fetal viability because of possible uterine stresses. Alth ough glycolysis predominates in the placenta. " the hypothesis th at other aerobic pathways in stressed placentas may produce lar ger amounts of adenosine triphosph ate cannot be excluded. The role of adenosine triphosphate in placental tissue has not yet been totall y characterized, although it has been shown to be important in the act of ion transport in placental tissue! ' and its levels have been investigated in preeclamptic placentas.": 13 Without kn owing the actual contribu tion of blood in an individual placenta we are un able to d etermine the true increase in oxidative phosphorylati on from the tissue per se. T o our knowled ge there are no accurate means to determine blood content with our methodology, and other mean s, such as histologic examination of each placenta to compare the intervillus vascular space with the villus tissue space, are at best estim ate s of the blood contribution.
552 Kay at al.
August 1992
Am
The most interesting and particularly remarkable finding in this study is the result that the overall phosphorus metabolite levels not restricted to adenosine triphosphate levels alone are higher in the growth-retarded placentas compared with the normal placentas (Table I). On the basis of our analysis there appear to be statistically significant differences in the metabolites of the growth-retarded placentas compared with the others in all categories, except for inorganic phosphorus and for glycerophosphorylethanolamine, where the concentration difference was close to being statistically significant with a p value of 0.0528. A statistically significant difference in the sugar phosphate levels, which includes such compounds as glucose-6-phosphate and fructose-f-phosphate, could represent enhanced carbohydrate metabolism, through glycolysis, in the growth-retarded placentas. The difference in the diphosphoglycerate levels is striking and, as previously mentioned, may be consistent with a higher amount of blood in the placenta. The differences noted in the glycerophosphorylethanolamine and glycerophosphorylcholine levels suggest a higher level of membrane phospholipid synthesis or turnover, because these are important components in cell membrane phospholipids. I' A previous study by Noyszewski et al." also reported an increase in the glycerophosphorylcholine peak intensity in growth-retarded placentas compared with normal placentas, although their results were not quantitated and were performed on pieces of fresh, nonextracted placenta bathed in a perfusate. In this current study the higher levels of nucleotide, carbohydrate, and membrane components in the growthretarded placentas may be important indicators of increased cell damage and repair in these placentas, compared with a normal healthy placenta. The highest levels of metabolites were found in inorganic phosphorus. Although there was no significant difference among the levels in the three groups, it was the highest in the IUCR group. which may be consistent with increased cellular metabolism because it has been suggested that large amounts of inorganic phosphate may be necessary for enzymatic function." Although the current results do not necessarily explain the cause of the growth retardation in the pregnancies studied, they do suggest that there may be placental response to hypoxia. Fox" and MacLennan et al. 18 have shown that human placental villi and trophoblasts in culture undergo syncytial degeneration and a marked increase in the number of cytotrophoblastic cells when exposed to hypoxic conditions. The higher levels of phosphorus-containing metabolites seen in our IUCR placentas could similarly represent repair and regenerative processes in the placenta in response to hypoxia. In these cases of idiopathic IUCR, hypoxia may exist in the placental bed but may not be recog-
J Obstet Gynecol
nized clinically. A placental reaction to hypoxia in these idiopathic cases of fetal growth retardation is consistent with other conditions known to be associated with IUCR, such as hypertension or infection, because these conditions can lead to h ypoxia in the placental bed. Therapeutic efforts directed at correcting the hypoxia, such as with the use of aspirin to decrease thromboxane production," may be further supported by these findings. Finally, it should be pointed out that the analyses undertaken in this study could have been carried out with other traditional biochemical tools. However, the NMR spectroscopic analyses offers the distinct advantage of a simultaneous study of all phosphorus-containing metabolites in one spectrum and the prospect for using NMR as an in vivo biochemical probe. Our current spectral results, which demonstrate an overall increase in phosphorus metabolites. have allowed us the insight that clinically unrecognized hypoxia may be one of the underlying causes of idiopathic IUCR. NMR spectra were obtained at the Duke University NMR spectroscopy center. which is funded by the National Institutes of Health, the National Science Foundation, the North Carolina Biotechnology Center, and Duke University. REFERENCES 1. Gabbe SG. Intrauterine growth retardation. In: Gabbe SG. Nieb yl JR, Simpson JL. eds. Obstetrics. New York : Churchill Livingstone, 1986 :769·71. 2. Rushton 01. Pathology of placenta. In : Wigglesworth JS, Singer DB, eds. Textbook of fetal and perinatal pathology. Boston: Blackwell Scientific, 1991 :186. 3. Nitzan M, Orloff S, Chrazanowska Bt., et at. Intrauterine growth retardation in renal insufficiency: an experimental model in the rat. AMJ OBSTET GYNECOL 1979; 133:40-3. 4. Rosso P. Maternal-fetal exchange during protein malnutrition in the rat: placental transfer of alpha-aminoisobutyric acid. J N utr 1977 ;107:2002-5. 5. Rosso P. Maternal-fetal exchange during protein malnutrition in the rat: placental transfer of glucose and a nonmetabolizable glucose analog. J Nutr 1977; 107:200610. 6. Kay HH, Gordon JD, Ribeiro AA , Spicer LD. Phosphorus-31 magnetic resonance spectroscopy of human placenta and quantitation with perchloric acid extracts. AM J OBSTET CYNECOL 1991 ;164 :80-7 . 7. Brenner WE, Edelman DA, Hendricks CH. A standard of fetal growth for the United States of America. AM J OBSTET CYNECOL 1976 ;126:555-64. 8. GardJ , AckermanJA. A 31p NMR external reference for intact biological systems.J Mag Res 1983;51:124-7. 9. Bottomley PA. Human in vivo NMR spectroscopy in diagnostic medicine: clinical tool or research probe? Radiology 1989 ;170 :1-15. 10. Bloxam DL. Human placental energy metabolism: its relevance to in vitro perfusion. Contrib Gynecol Obstet 1985; 13:59-69. 11. Miller RK, Berndt WO o Mechanisms of transport across the placenta: an in vitro approach. Life Sci 1975;16:7-30. 12. Hoover CR, Bartholomew RA. Acid-soluble nucleotides in placentae from normal and toxic patients. Obstet CynecoI1959;14:309-21. 13. Bloxam DL, Bullen BE, Walters BNJ, Lao TT. Placental
Volume 167 Number 2
glycolysis and energy metabolism in preeclampsia. AM] OBSTET GYNECOL 1987;157:97-101. 14. Sostman HD, Armitage 1M, Fischer ]J. NMR in cancer. 1. High resolution spectroscopy of tumors. Mag Resonance Imaging 1984;2:265-78. 15. Noyszewski EA, Raman], Trupin SR, McFarlin BL, Dawson MJ. Phosphorus 31 nuclear magnetic resonance examination of female reproductive tissues. AM ] OBSTET GYNECOL 1989;161:282-8. 16. lies RA, Stevens AN, Griffiths]R, Morris PG. Phosphorylation status of liver by 3!P NMR spectroscopy, and its implications for metabolic control. Biochem] 1985;229: 141-51.
NMR spectroscopy of normal and growth-retarded placentas
17. Fox H. Effect of hypoxia on trophoblast in organ culture. AM] OBSTET GYNECOL 1970; 107: 1058-64. 18. MacLennan AH, Sharp F, Shaw-Dunn J. The ultrastructure of human trophoblast in spontaneous and induced hypoxia using a system of organ culture. ] Obstet Gynaecol Br Commonw 1972;79:113-21. 19. Trudinger B], Cook CM, Thompson RS, Giles WB, Connelly A. Low-dose aspirin therapy improves fetal weight in umbilical placental insufficiency. AM] OBSTET GYNECOL 1988;159:681-5.
Electrocortical activity, electroocular activity, and breathing movements in fetal sheep with prolonged and graded hypoxemia Bryan S. Richardson, MD, Lesley Carmichael, BSc, Jacobus Homan, BSc, and John E. Patrick, MD London, Ontario, Canada OBJECTIVE: Our objective was to determine the effect of a prolonged and graded reduction in fetal arterial oxygen saturation on electrocortical activity and associated biophysical variables. STUDY DESIGN: Fourteen unanesthetized fetal sheep were studied between 126 and 135 days' gestation with continuous monitoring of electrocortical and electroocular activity and breathing movements, during a 24-hour control period, and subsequently during 4 days of prolonged and graded hypoxemia induced by progressively lowering the maternal inspired oxygen concentration. RESULTS: Graded reduction in fetal arterial oxygen saturation resulted in little change in arterial pH until close to 30% when metabolic acidemia was apparent. The incidence of low-voltage electrocortical activity, electroocular activity, and breathing movements were marginally decreased with hypoxemia alone; however, a significant decrease was not apparent until associated with the onset of fetal acidemia. CONCLUSION: Hypoxemia of a chronic nature must approach the level at which acidemia becomes apparent before a marked change in fetal behavioral activity is noted. (AM J OSSTET GVNECOL 1992;167:553-8.)
Key words: Fetal behavior state, breathing movements, hypoxemia It is well recognized that the human fetus near term displays a periodicity in behavioral and biophysical parameters with coincidence of such into well-defined behavioral states similar to those described for the human newborn.' Cyclic changes in fetal behavioral state and From the Department of Obstetrics and Gynaecology and Physiology, St. Joseph's Health Centre, Lawson Research Institute, The University of Western Ontario. Supported by grants from the Canadian Medical Research Council. Received for publication October 28,1991; revised March 13,1992; accepted March 31,1992. Reprint requests:Bryan S. Richardson, MD, Department ofObstetrics and Gynaecology, Lawson Research Institute, St. Joseph's Hospital, 268 Grosvenor St., London, Ontario, Canada N6A 4V2.
6/1/38218
state-related parameters are thus characteristic of the healthy fetus and provide a basis for the biophysical assessment of fetal health." However, although biophysical parameters appear to reflect fetal health at the time of study; their time course for alteration when fetal oxygenation becomes restricted is unknown, which may limit their usefulness in predicting fetal compromise. Studies with the chronically catheterized fetal lamb with short-term, induced hypoxemia show the incidence of low-voltage electrocortical activity, electroocular activity, and breathing movements to be variably reduced.v" Although further supporting the role ofbiophysical parameters in the assessment of fetal health,
553
Key words: Phosphorus 31 nuclear magnetic resonance spectroscopy, human placenta, perchloric acid extraction, intrauterine growth retardation Intrauterine growth retardation (IUGR) is a complication of pregnancy affecting as many as 4% to 7% of all pregnancies.' The causes of fetal growth retardation include maternal hypertension, cigarette and other substance abuse, in ut ero infection, and chromosomal disorders. In many cases, however, a specific cause is never identified. The placenta, which serves many critical functions for the fetus, is an important organ to study in any attempt to understand the cause of idiopathic IUGR. Very little is known about the role of the placenta in fetal IUGR. Most studies have focused on pathologic and histologic examinations. These investigations, however, tended to be nonspecific because of the many different causes of IUGR.2 Other studies on animals have reported a decrease in glucose and amino acid transfer to the fetus. "" Few stud ies, however, ha ve yet focused on a biochemical analysisof placental metabolic fun ction. Such analysis may be particularly informative in cases where the cause is unidentified.
From the Departments of Obstetrics and Gynecology: Biochemistry,' and R adiology: Duke University M edical Center. Supp orted by National Institutes of Health grant N o. 1K08HD00822-02 (H. K. ). R eceioedfor publication August 4, 1991; revisedJanuary 15,1992; accepted March 31,1992. R eprint requests: Helen H. Kay, MD, Box 3373, Duke University M edical Center, Durham , NC 27710. 611138320
548
Nuclear magnetic resonance (NMR ) spectroscopy offer s a unique method for studying placental metabolic fun ction. It is a noninvasive technology that simultaneously monitors naturally occurring molecules in biologic tissue . In a previous report" we described the phosphorus 31 magnetic reson ance spectra of fresh human placental tissues and perchloric acid-extracted tissue. We also detailed the quantitative analysis of adenosine triphosphate and inorganic phosphorus levels in the placental extracts. In the current study we extend our quantitative analy sis to other metabolites easily visible in the phosphorus spectra, such as sugar phosphate, diphosphoglycerate, glycerophosphorylethanolamine, and glycerophosphorylcholine. We have focused our anal ysis on a compa rison of placentas from idiopathic growth-retarded pregnancies and normal pregnancies to gain further insights into the cause of the growth retardation.
Material and methods Five placentas from pregnancies antenatally identified as complicated by IUGR of unknown cause were obtained at the time of delivery by cesarean section. T hese were confirmed as growth-retarded placentas by neonatal weights < 10th percentile with Brenner's growth curve.' The mean birth weight was 1253 ± 207 gm at a mean gestational age of 35 .3 ± 0.9 weeks con firmed by ultrasonographic examination before 20 weeks . Thorough chart review and patient interview
NMR spectroscopy of normal and growth-retarded placentas
Volume 167 Nu mb er 2
revealed no specific cause suc h as hypertension, cigarette or drug abuse, viral illnesses, or prenatal complications , i.e., bleeding. The mean placental weight was 198 :t 32 gm, and there were no gross abnormalities such as infarctions or abruptio placentae seen on inspection. Small pieces of placental tissue were removed from multiple regions on th e mat ernal side by sharp dissection and placed into liquid nitrogen within I min ute a nd sto red in a-70° C free zer for later anal ysis. Similarl y, placentas from normal, un complicated pregnan cies obtained from eight vaginal deli veries and eight elective cesarean section deliver ies were harvested and fr ozen in the same fashion . Placentas in all three groups were obtained in the third tr imester. At the time of extraction the tissue was removed from the freezer and pulverized in liquid nitrogen with a stainless steel Waring blender (Dynamics Corp. of America, New Hartford, Conn .). A perchloric acid extract was performed as pr eviou sly des cribed." Briefly, a 20 gm sample of finely gro und placental tissue was mixed with 20 ml of I mol / L perchloric acid that had been precooled to -78 C in dry ice. The perchloric acid - tissue mixture was allowed to warm to approximatel y - 10° C in the ambi en t air for 15 minutes during which time the metabolites wer e extr acted . The extract was then centrifuged at 40 ,000g at _4° C for 20 minute s to separate the solid fr action from the liquid fra ction . T he liquid supernatant conta ining the extracted metabolit es was then decanted into a chilled beaker a nd neutralized to pH 7.0 at 4° C through the dropwise add ition of potass ium hydroxide. The potassium perchlorate precipitated durin g neutralization was re moved by a second centrifugation step identical to the first, and the supernatant was then passed over a Chelex- l 00 chelating resin (Sigma Chemical Co., St Loui s) to remove divalent cations. The extract was again titr at ed to neutral pH with diluted perchloric acid, frozen in a vacuu m flask (Virtis Co ., Gardiner, N.Y.), and lyophilized to dryness. The extract was then resuspended in 5 ml of 40 % deuterated water for quantitat ion of spectra. The extracts were placed in the outer chamber of a coaxial 10 mm NMR tube (Wilmad Glass, Buen a, N.J.) with an external reference standard of 25 mmol / L hexachlorocyclotriphosphazene mixed with 3 1 mmol /L chro mium acet ylaceto nate (Ald rich Chemical lnc., Milwaukee)" in an inn er coaxial insert. Spectra were obtained with a 21 usee pul se giving a 70° flip angle, an acquisition period of 1.12 seconds, a relaxation dela y of 15 seconds (total pulse recycle time 16.12 seconds), and 160 acqui sition s. When phosphorus nuclei have su fficient time to return to their equilibrium state , i.e., full y relaxed state , their peak intensities are proportional to the concentration of the metabolite present. To assess the consequence of the pulse recycle time (acquisition time of 0
549
1.12 seconds plus vari able del ay) o n the extent of nuclear relaxation, spectra were obtained at rec ycle times of6 .12, 11.12 , 16.12 ,21.12, 31.1 2, and 61.12 seconds. T he hexachlorocyclotriphosphazene and (3-adenosine triphosph ate resonances are fully relaxed under con dition s of our "sta ndard" 16.12-second recycle time." The suga r phosphate, ino rgani c phosphate, glyceroph osphorylethanolamine, and glycerophosphorylcholine resonances, respectively, return to 93.7%, 97 .4%, 95 .1%, and 98 .6% of their equ ilibrium states at the 16.12-second recycle time . The following "satura tion correction factors " are then used to correct the integra ted intensities of the signals obtained at 16.12 seconds: adenosine triphosphate 1.00, hexachlorocyclotriphosphazene 1.00; sugar ph osphate 1.07, inorganic ph osphate 1.03; glycerophosphorylcholine 1.05, and glycero phosphorylethano lamine 1.0I before calculation of metabolite concentrations . T he quantitative anal ysis is previou sly reported." Metabolite quantitation was performed with the external re feren ce standard of hex achlo rocyclotriphosphazene and chro miu m acet ylaceton ate as a spin relaxation agent (Ald rich) in a coaxial capillary system. In br ief, the integ rated peak are a of a ph osphorus resonance was divided by the integrated peak area of the hexachloroc yc1otriphosphazene reference signal. The rati o for th e (3-adenosine triphosphate re sonance is then used to qu antitate the adenosine triphosphate metabolite present with a calibration cu rve previously constr ucted for varying known concen tra tions of adenosine triphosphate. The ad en osine tr iphosphate levels and the (3-adenosine triphosph ate /hexachlorocyclotriph osphazene ratio are related by the equation }' = 0.035 + O.088X. After calculation of the concentr ation of adenosine triphosphate, the remaining peak ratios and their corresponding satu ration correction factor s were then used to determine the final concentr ation of the other metabolites. All spe ctr a were obtained on a 300 MHz NMR spectrometer (model GN-300 WB General Electric , Fre mont, Calif.) interfaced with a wide-bore superconducting magnet (89 mrn bore, 7.0 tesla field) in the Duke Magnetic Resonance Spectroscopy Center. A standa rd 20 mm broad band probe (General Electric) tu ned to 121.4 MHz was used for SI p spectroscopy. Spectroscopic data were collected at the ambient op erating probe temperature (approximately 22° C) without temperature regulation. T o avoid radiofrequency heating effects on the tissue a nd extracts, no proton decoupling was used . :lI p chemical shifts are referenced to phosphocreatine at 0.00 parts /m illion. Stat istical analysis was accompli shed with a Statgraphics program (Statistical Graphics Corporation, STSC, Rockville, Md.). The means and standard deviation of each metabolite were calculated for the three
550 Kayet al.
August 1992
Am
J Obstet Cynecol
Pi
HCCTP
SP
GPC
DPG Y-ATP
a-ATP
B-ATP
I -20
I'
10
-30
PPM
Fig. 1. Phosphorus NMR spectrum of perchloric acid-extracted placental tissue from normal vaginal delivery. Note resonances for external reference standard hexachlorocyclotriphosphazene (HCCTP), diphosphoglycerate (DPG), sugar phosphate (SP), inorganic phosphate (Pi), glycerophosphorylethanolamine (GPE), glycerophosphorylcholine (GPC), phosphocreatine (PCr), and adenosine triphosphate (A TP). groups (vaginal delivery, cesarean section, and growth retardation). A one-way analysis of variance was performed to determine the presence of a statistical difference among the means of the three groups. A multiple-range analysis or a t test was performed to determine specific differences among each of the three groups. Results
Fig. I shows the phosphorus spectrum from a perchloric acid extract of a normal placenta. Of note are the distinct and well resolved resonances for hexachlorocyclotriphosphazene, sugar phosphate, diphosphoglycerate, inorganic phosphorus, glycerophosphorylethanolamine, glycerophosphorylcholine, and the three phosphorus nuclei of adenosine triphosphate. Note that there is very little phosphocreatine at 0 parts/million, and there is a large amount of inorganic phosphorus. Table I lists the placental metabolite concentrations determined by 3lp NMR for vaginal-delivery placentas, cesarean-section placentas and growth-retarded placentas delivered by cesarean section. The table specifically lists the mean concentration and standard deviation for each of the metabolites of the three placenta
groups. The mean concentrations of the phosphorus metabolites are highest in the IUGR group, by almost a factor of two compared with the vaginal and cesarean section groups, except for inorganic phosphorus. There are no statistically significant differences among the metabolites between the vaginal and cesarean section placentas. However, there are statistically significant differences among the adenosine triphosphate, sugar phosphate, diphosphoglycerate, and glycerophosphorylcholine levels between the growth-retarded placentas and those from normal vaginal and cesarean section deliveries. Comment
The current study is a comparative analysis of placentas from normal and abnormal pregnancies with phosphorus NMR spectroscopy. Our findings suggest that NMR spectroscopy, with its ability to simultaneously display the levels of several metabolites, can be extremely useful for in vitro and in vivo biochemical analysis. In this study the NMR analysis yields further insights into placental metabolism. Human in vivo NMR studies in diagnostic medicine are currently focused on ischemic brain injury, muscle dystrophy, and tumor response to chemotherapy." With continued de-
Volum e 16i Number 2
NMR spectroscopy of normal and growth-retarded placentas 551
Table I. Placental metabolite concentrations determined by phosphorus NMR spe ctroscopy
Metabolit e
Adenosine triphosphate Sugar phosphate Inorgan ic phosphate Diph osphoglycerate Glycerophosphoryleth an olamine Glycerophosphorylcholin e
No rmal vaginal delivery (n = 8) (mean ± S D) (um ol! gm wet weight)
0.270 1.l03 2.586 0.383 0.273 0.556
± 0.189 ± 0.308 ± 0.308
± 0.135 ± 0.095 ± 0 .139
No rmal cesarean section (n = 8) (mean ± S D) (umol ]gm wet weight)
0.426 1.502 3.477 0.4 26 0.347 0.685
± 0.154
± 0.305+
± 1.643
± 0.151+
± 0.084
± 0. 130:j:
Growth retardation (n = 5) (mean ± SD) (u mol! gm wet weight)
0.547 2.137 3.720 0.712 0.440 1.112
± 0.160t
0.31 9§ 1.777 0.090§ 0.16 6 ± 0.348§
± ± ± ±
p Valu e*
0.0295 0.0001 0.2740 0.001 0 0.0528 0.0005
*T he p values obtained fro m a one-way ana lysis of vari an ce to test differ en ces between means in all three gro ups . t Statistically significant difference between the growth-retarded and normal vaginal d eliver y groups by a t test but not th e multiple range test. :j:Statistically significant di fference betwee n the growth-retarded and normal cesarean section groups by mul tiple range test. §Statistically significant difference between the growth-retarded and normal vaginal deliv er y gro ups by multiple range test.
velopments in technology, detailed an alysis ma y someda y be feasible for in vivo spectrosco pic studies of the human placenta. Our current report focu ses on differences bet wee n spectra fr om tissue extracted with perchloric acid both for th e increased resolution available to identify biochemical change and with the expectation that the se differences, once established, may be visible in whole placental tissue. It is an important first step to characterize th e NMR spectra of human placental extracts to facilitate the analy sis of peaks in the spectra from gross tissue for relevant biochemical and diagnostic studies in the future . Our previous preliminary studies on gro ss placental tissue" demonstrated obvious resonances for sugar phosph ate, inorganic phosphorus and adenosine triphosphate. The diphosphoglycerate, glycerophosphorylethanolam ine , and glycerophosphorykholine resonances were not well resolved. In addition, phosphocreatine was absent in the spectrum of gross tissue. However, with perchloric acid extraction we were able to con firm the presence of phosphocreatine albeit in very small qu antities. We were also able to determine tissue ad en osine triphosph ate levels from the extracts. In this stu dy we performed further analyses on the spectra from perchlori c acid-extracted tissue . We found different ad enosine triphosphate levels in the three groups studied. They were lowest in the vaginal delivery group, although not statistically different from the cesarean section gr oup. The adenosine triphosphate level is highest in the IUGR group with a statistical difference from the vaginal delivery group. Bloxam 10 also reported lower levels of adenosine triphosphat e in placentas from vaginal deliveries than from placentas from cesarean section. This may, in part, be due to varying a mounts of blood within the placentas. Less blood is expected in vaginal delivery placentas because blood is squeezed out of the sinuses during labor and delivery. We have previously found"
that the blood concentration within th e placentas is significant and may accou nt for as mu ch as 30% to 50% of the total amou nt of ade nosine triphosphate seen. Therefore the fact that the highest level was found in the I UGR placentas is interesting and suggests that these placentas may have th e largest amount of blood . This is further supported by the fact th at diphosphoglycera te , which is a major component of blood, is present at th e highest level in the IUGR gr oup and is statistically di fferent from the other two groups. This increase in blood content in the IUGR placentas may represent a com pensatory increase in blood flow. It is also possible th at differences in the adenosine triphosphate levels re sult from varying oxidative phosphorylation requirements within the placentas at the time of delivery. Levels are likely to be lower in the vaginal delivery group because the adenosine triphosphate is used during the process of labor. T he increased levels in placentas from growth-retarded pregnancies, on the other hand, could arise from a need to increase oxidative phosphorylation to maintain fetal viability because of possible uterine stresses. Alth ough glycolysis predominates in the placenta. " the hypothesis th at other aerobic pathways in stressed placentas may produce lar ger amounts of adenosine triphosph ate cannot be excluded. The role of adenosine triphosphate in placental tissue has not yet been totall y characterized, although it has been shown to be important in the act of ion transport in placental tissue! ' and its levels have been investigated in preeclamptic placentas.": 13 Without kn owing the actual contribu tion of blood in an individual placenta we are un able to d etermine the true increase in oxidative phosphorylati on from the tissue per se. T o our knowled ge there are no accurate means to determine blood content with our methodology, and other mean s, such as histologic examination of each placenta to compare the intervillus vascular space with the villus tissue space, are at best estim ate s of the blood contribution.
552 Kay at al.
August 1992
Am
The most interesting and particularly remarkable finding in this study is the result that the overall phosphorus metabolite levels not restricted to adenosine triphosphate levels alone are higher in the growth-retarded placentas compared with the normal placentas (Table I). On the basis of our analysis there appear to be statistically significant differences in the metabolites of the growth-retarded placentas compared with the others in all categories, except for inorganic phosphorus and for glycerophosphorylethanolamine, where the concentration difference was close to being statistically significant with a p value of 0.0528. A statistically significant difference in the sugar phosphate levels, which includes such compounds as glucose-6-phosphate and fructose-f-phosphate, could represent enhanced carbohydrate metabolism, through glycolysis, in the growth-retarded placentas. The difference in the diphosphoglycerate levels is striking and, as previously mentioned, may be consistent with a higher amount of blood in the placenta. The differences noted in the glycerophosphorylethanolamine and glycerophosphorylcholine levels suggest a higher level of membrane phospholipid synthesis or turnover, because these are important components in cell membrane phospholipids. I' A previous study by Noyszewski et al." also reported an increase in the glycerophosphorylcholine peak intensity in growth-retarded placentas compared with normal placentas, although their results were not quantitated and were performed on pieces of fresh, nonextracted placenta bathed in a perfusate. In this current study the higher levels of nucleotide, carbohydrate, and membrane components in the growthretarded placentas may be important indicators of increased cell damage and repair in these placentas, compared with a normal healthy placenta. The highest levels of metabolites were found in inorganic phosphorus. Although there was no significant difference among the levels in the three groups, it was the highest in the IUCR group. which may be consistent with increased cellular metabolism because it has been suggested that large amounts of inorganic phosphate may be necessary for enzymatic function." Although the current results do not necessarily explain the cause of the growth retardation in the pregnancies studied, they do suggest that there may be placental response to hypoxia. Fox" and MacLennan et al. 18 have shown that human placental villi and trophoblasts in culture undergo syncytial degeneration and a marked increase in the number of cytotrophoblastic cells when exposed to hypoxic conditions. The higher levels of phosphorus-containing metabolites seen in our IUCR placentas could similarly represent repair and regenerative processes in the placenta in response to hypoxia. In these cases of idiopathic IUCR, hypoxia may exist in the placental bed but may not be recog-
J Obstet Gynecol
nized clinically. A placental reaction to hypoxia in these idiopathic cases of fetal growth retardation is consistent with other conditions known to be associated with IUCR, such as hypertension or infection, because these conditions can lead to h ypoxia in the placental bed. Therapeutic efforts directed at correcting the hypoxia, such as with the use of aspirin to decrease thromboxane production," may be further supported by these findings. Finally, it should be pointed out that the analyses undertaken in this study could have been carried out with other traditional biochemical tools. However, the NMR spectroscopic analyses offers the distinct advantage of a simultaneous study of all phosphorus-containing metabolites in one spectrum and the prospect for using NMR as an in vivo biochemical probe. Our current spectral results, which demonstrate an overall increase in phosphorus metabolites. have allowed us the insight that clinically unrecognized hypoxia may be one of the underlying causes of idiopathic IUCR. NMR spectra were obtained at the Duke University NMR spectroscopy center. which is funded by the National Institutes of Health, the National Science Foundation, the North Carolina Biotechnology Center, and Duke University. REFERENCES 1. Gabbe SG. Intrauterine growth retardation. In: Gabbe SG. Nieb yl JR, Simpson JL. eds. Obstetrics. New York : Churchill Livingstone, 1986 :769·71. 2. Rushton 01. Pathology of placenta. In : Wigglesworth JS, Singer DB, eds. Textbook of fetal and perinatal pathology. Boston: Blackwell Scientific, 1991 :186. 3. Nitzan M, Orloff S, Chrazanowska Bt., et at. Intrauterine growth retardation in renal insufficiency: an experimental model in the rat. AMJ OBSTET GYNECOL 1979; 133:40-3. 4. Rosso P. Maternal-fetal exchange during protein malnutrition in the rat: placental transfer of alpha-aminoisobutyric acid. J N utr 1977 ;107:2002-5. 5. Rosso P. Maternal-fetal exchange during protein malnutrition in the rat: placental transfer of glucose and a nonmetabolizable glucose analog. J Nutr 1977; 107:200610. 6. Kay HH, Gordon JD, Ribeiro AA , Spicer LD. Phosphorus-31 magnetic resonance spectroscopy of human placenta and quantitation with perchloric acid extracts. AM J OBSTET CYNECOL 1991 ;164 :80-7 . 7. Brenner WE, Edelman DA, Hendricks CH. A standard of fetal growth for the United States of America. AM J OBSTET CYNECOL 1976 ;126:555-64. 8. GardJ , AckermanJA. A 31p NMR external reference for intact biological systems.J Mag Res 1983;51:124-7. 9. Bottomley PA. Human in vivo NMR spectroscopy in diagnostic medicine: clinical tool or research probe? Radiology 1989 ;170 :1-15. 10. Bloxam DL. Human placental energy metabolism: its relevance to in vitro perfusion. Contrib Gynecol Obstet 1985; 13:59-69. 11. Miller RK, Berndt WO o Mechanisms of transport across the placenta: an in vitro approach. Life Sci 1975;16:7-30. 12. Hoover CR, Bartholomew RA. Acid-soluble nucleotides in placentae from normal and toxic patients. Obstet CynecoI1959;14:309-21. 13. Bloxam DL, Bullen BE, Walters BNJ, Lao TT. Placental
Volume 167 Number 2
glycolysis and energy metabolism in preeclampsia. AM] OBSTET GYNECOL 1987;157:97-101. 14. Sostman HD, Armitage 1M, Fischer ]J. NMR in cancer. 1. High resolution spectroscopy of tumors. Mag Resonance Imaging 1984;2:265-78. 15. Noyszewski EA, Raman], Trupin SR, McFarlin BL, Dawson MJ. Phosphorus 31 nuclear magnetic resonance examination of female reproductive tissues. AM ] OBSTET GYNECOL 1989;161:282-8. 16. lies RA, Stevens AN, Griffiths]R, Morris PG. Phosphorylation status of liver by 3!P NMR spectroscopy, and its implications for metabolic control. Biochem] 1985;229: 141-51.
NMR spectroscopy of normal and growth-retarded placentas
17. Fox H. Effect of hypoxia on trophoblast in organ culture. AM] OBSTET GYNECOL 1970; 107: 1058-64. 18. MacLennan AH, Sharp F, Shaw-Dunn J. The ultrastructure of human trophoblast in spontaneous and induced hypoxia using a system of organ culture. ] Obstet Gynaecol Br Commonw 1972;79:113-21. 19. Trudinger B], Cook CM, Thompson RS, Giles WB, Connelly A. Low-dose aspirin therapy improves fetal weight in umbilical placental insufficiency. AM] OBSTET GYNECOL 1988;159:681-5.
Electrocortical activity, electroocular activity, and breathing movements in fetal sheep with prolonged and graded hypoxemia Bryan S. Richardson, MD, Lesley Carmichael, BSc, Jacobus Homan, BSc, and John E. Patrick, MD London, Ontario, Canada OBJECTIVE: Our objective was to determine the effect of a prolonged and graded reduction in fetal arterial oxygen saturation on electrocortical activity and associated biophysical variables. STUDY DESIGN: Fourteen unanesthetized fetal sheep were studied between 126 and 135 days' gestation with continuous monitoring of electrocortical and electroocular activity and breathing movements, during a 24-hour control period, and subsequently during 4 days of prolonged and graded hypoxemia induced by progressively lowering the maternal inspired oxygen concentration. RESULTS: Graded reduction in fetal arterial oxygen saturation resulted in little change in arterial pH until close to 30% when metabolic acidemia was apparent. The incidence of low-voltage electrocortical activity, electroocular activity, and breathing movements were marginally decreased with hypoxemia alone; however, a significant decrease was not apparent until associated with the onset of fetal acidemia. CONCLUSION: Hypoxemia of a chronic nature must approach the level at which acidemia becomes apparent before a marked change in fetal behavioral activity is noted. (AM J OSSTET GVNECOL 1992;167:553-8.)
Key words: Fetal behavior state, breathing movements, hypoxemia It is well recognized that the human fetus near term displays a periodicity in behavioral and biophysical parameters with coincidence of such into well-defined behavioral states similar to those described for the human newborn.' Cyclic changes in fetal behavioral state and From the Department of Obstetrics and Gynaecology and Physiology, St. Joseph's Health Centre, Lawson Research Institute, The University of Western Ontario. Supported by grants from the Canadian Medical Research Council. Received for publication October 28,1991; revised March 13,1992; accepted March 31,1992. Reprint requests:Bryan S. Richardson, MD, Department ofObstetrics and Gynaecology, Lawson Research Institute, St. Joseph's Hospital, 268 Grosvenor St., London, Ontario, Canada N6A 4V2.
6/1/38218
state-related parameters are thus characteristic of the healthy fetus and provide a basis for the biophysical assessment of fetal health." However, although biophysical parameters appear to reflect fetal health at the time of study; their time course for alteration when fetal oxygenation becomes restricted is unknown, which may limit their usefulness in predicting fetal compromise. Studies with the chronically catheterized fetal lamb with short-term, induced hypoxemia show the incidence of low-voltage electrocortical activity, electroocular activity, and breathing movements to be variably reduced.v" Although further supporting the role ofbiophysical parameters in the assessment of fetal health,
553
