Oral Concurrent Session F OB Anesthesia and Pharmacology Medical Complications of Pregnancy

Oral Concurrent Session F OB Anesthesia and Pharmacology Medical Complications of Pregnancy Friday, January 28, 1994 1:00 p.m. - 3:00 p.m. Moderator: Richard H. Paul, MD Past President, SPO

Ballroom C Abstract Numbers 61 - 68

SPO Abstracts

Volume 170, Number 1, Part 2

Am J Obstet Gynecol

61

LOW-DOSE ASPIRIN IN NULliPAROUS WOMEN: SAFETY OF EPIDURAL AND CORRELATION BETWEEN BLEEDING TIME AND MATERNAL-NEONATAL BLEEDING COMPUCATIONS. a...siIlID, S. Caritis, E. Thom, X K. Shaw, D. McNe16s, X and the NICHD MFM Network, Bethesda, Maryland. . OBJECTIVE: To determine the frequency and safety of the use of epidural anesthesia and the correlation between bleeding time and maternal-neonatal bleeding complications in a group of pregnant women who participated in a muhicenter lowdose aspirin in pregnancy trial. STUDY DESIGN: 2985 nulliparous women from 7 centers were randomly assigned at t3-27 weeks to receive 60 mg/d.of aspirin or a placebo. Data regarding type of anesthesia used for labor and delivery were collected in 5 centers who enrolled 2579 women. Knowledge re9arding use of epidural was known in 1629 (63%) of the women: 891 (55%) receIVed epidural anesthesia. RESULTS: Among the 891 women known to have received epidural, 451 received low-dose aspirin and 440 received a placebo. The frequency of epidural use among the 5 centers ranged from 27% to 94%. There was no instance of bleeding related to epidural use. One of the 5 centers also obtained bleeding times in 303 women (149 received aspirin and 154 received placebo). The mean±SD bleeding time in WlTSl receiving low-10 minutes was higher in the aspirin group (14.1% vs 5.2%, p=O.Ol). Interestingly, women who received epidural had a lower bleeding time than those not receiving epidural (p=0.003) irrespective of the treatment used. Despite the difference in bleeding time between aspirin and placebo, we found no differences in maternal and neonatal bleeding complications (Table). No baby had excessive bleeding from circumcision. Aspirin (n=149)

Placebo (n=154)

6 (4.0)

7 (4.5) 546±163 4.43±3.2 4 (2.5)

Postpartum hemorrhage #(%) Estimated blood loss (ml)' Change in hematocrit (%) Neonatal bleeding #(%)

555±196 4.43±3.5 3 (2.0)

• Power of 99.7% to detect a difference of 100 m!. CONCLUSIONS: Epidural anesthesia is safe to use in women receiving low-dose aspirin in pregnancy. Despite an increased bleeding time in pregnant women receiving low-dose aspirin, maternal-neonatal bleeding complications were not increased.

62

EFFECT OF ASPIRtN AND tNDOMETHACIN ON PROSTACYCUN AND THROMBOXANE PRODUCTION BY PLACENTAL TISSUE INCUBATED WITH IgG FRACTIONS FROM PATIENTS WITH LUPUS ANTICOAGULANT. AM Peaceman. K Rehnberg'. Department of Ob/Gyn, Northwestern University Medical School, Chicago II. OBJECTIVE: We assessed the effect of aspirin (ASA) and indomethacin (I) on prostanoid production by placentat tissue incubated with IgG fractions from patients with lupus anticoagulant (LA). METHODS: IgG fractions ware prepared using an ammonium sulfate precipitation method from the plasma of 6 non-pregnant patients with the antiphospholipid antibody (APA) syndrome and demonstrable LA. Fresh placentas were obtained from norma.1 term pregnanCies, and 350 mg

explants of ptacental tissue were incubated with 3 mg of IgG from each of the LA patients. For 8ach patjent, incubations were performed in duplicate with final concentrations of to·oM ASA. to-'M I. and no added nonsteroidal anti-inflammatory drug (NSAID). Aliquots were removed at intervals up to 48 hr of incubation and assessed for placental

prostacyclin (PGI) and thromboxane (TX) production by RIA of the stable metabolites PGF,. and TXB 2 • RESULTS: TX production by 20 hr of incubation is displayed in the figure.

A 49-

81 % reduction in placental TX production was seen with the addition of ASA (paired t-test. p < _01 L to levels previously seen with incubation of placental tissue alone. A similar decrease in TX production (34-76%, p<.05) c .., was noted with the addition of I. Decreased PGI production (26-46%. p < .01) was also : tOO observed with the addition of I, but ASA at 1-, ' - - - - - this dose did not affect PGt production_ -ASA CONCLUSION: Addition of both ASA and I at levels comparabla to

'"

therapeutic plasma levels eliminates the increased TX production seen

with placental tissue incubated with LA-lgG. These data provide a physiologic basis for treatment of APA related pregnancy loss with NSAID.

63

WHY CAN'T GOOD GLYCEMIC CONTROL OF ·DIABETES IN PREGNANCY ELIMINATE THE PROBLEM OF MACROSOMIA? B. Rosenn, M_ Miodovnik, G. Holcberg-, J. Khoury, R. Jaekle, T.A. Siddiqi. DepL Ob/Gyn; University of Cincinnati, Cincinnati OH. BACKGR.OUND: Although macrosomia in infants of diabetic women is assumed to be related to poor glycemic control during pregnancy, a .high rate of macrosomia persists in Type I (lDDM) diabetic pregnancies despite intensive insulin therapy. OBJECTIVE: To test the hypothesis that macrosomia in IDDM is associated with poor glycemic control during pregnancy_ DESIGN: 293 women with IDDM and 33 women with Type n (NIDDM, Maturity onset), diabetes who delivered viable singleton infants after· 34 weeks gestation were included in the analysis. All received intensive insulin therapy during pregnancy, and self-monitored blOod glucose concentrations (BG) 4-6 times daily. Gestational age was verified by early sonogram. Birth weight percentiles were determined from locat race and sex specific growth curves, and correlated with maternal glycemic control during pregnancy. RESULTS: Among infants ofIDDM women, 39% were LGA (>90th%) and 64% of these were >95th%. Glycohernoglobin concentrations (GHb) were within normal range in 74% of IDDM mothers by the second trimester. Third trimester mean (:!:SD) preprandial BG was 113:!:25 mg/dL, and postprandial BG was 134 :!:32mg/dL (interquartile ranges 95-125 and 11 0-151 respectively). The rate of LGA was related to weight gain in pregnancy, but was similar in women that had normat GHb (40%LGA) and in those that had higher GHb (36%LGA). LGA did not corretate with mean pre- or postprandial BG during any stage in pregnancy. In contrast, LGA in women with NIDDM was significantly associated with initial maternal weight and glucose control during pregnancy. CONCLUSION: Macrosomia is a persistent problem in IDDM pregnancies despite intensive insulin therapy and gOod glycemic control. We speculate that deficiency of counterregulatory hormonal responses in IDDM results in recurrent maternal hypoglycemic episodes with rebound excess of fOod intake, predisposing these infants to acceterated growth in utero. (Supported in part by NIH Grant HD 11725)

UNIVERSAL CRITERIA FOR GESTATIONAL DIABETES: 64 TOWARD THE 75 GRAM GLUCOSE TOLERANCE TEST IN PREGNANCY. D.Sacks. x

J. Greenspoon. S. Abu-Fadil • H. Henry, G. Wolde-~ J. Yao x • Depts. Ob/Gyn, Kaiser Foundation Hospital, Bellflower, and Cedars-Sinai Medical Ctr, los Angeles, Ca. OBJECTIVE: The Thi rd Internati ona 1 Workshop-Conference on gestational diabetes concluded that the 75 g. gliJcose tolerance test (GTT) was the test most likely to achieve international consensus to define gestational diabetes. The purpose of this study was to develop norms for that test in pregnancy, and to define glucose intolerance by the relationship between maternal glucose values and neonatal macrosomia. . STUDY DESIGN: 3063 unselected. predominantly hispanic gravidas who were not previously known to have diabetes were given a 75 g. 2-hour GTT. Fasting, one. and two hour samptes were drawn. Only those having a fasting plasma glucose (FPG) 2105 mg/dl and/or a 2-hour post-glucola value (2hrPG) 2200 mg/dl received diet and/or insul in therapy. Birthweights of only 1iveborn singletons whose mothers had FPG <105 mg/dl and 2hrPG <200 mg/dl were used to calculate relationships between glucose 1eve 1sand bi rthwei ghts. Macrosomi a was defi ned as a birthweight >90%ile for gestational age. RESULTS: The mean+2S.D. values for fasting. 1, and 2 hours were respectively 102 mg/dl.. 197' mg/dl.. and 160 mg/dl. In a multiple logistic regreSSion model, th.e factors found to be significantly associated with macrosomia and their odds ratios were maternal race (white/black 2.71; hispanic/black 1.90). parity (2.59). prepregnancy BMI (1.06), weight gain (4.20). gestational age at testing (0.95), FPG (1.02), and 1hrPG (1.01). There was no single glucose value or combination of values above which the risk for macrosomia was significantly increased. CONCLUSIONS: Within the range of GTT values studied: (1) Fetal macrosomia is influenced by factors other than maternal glucose levels. (2) There does not appear to be a threshhold point that distinguises women at increased risk for macrosomia.

293

294

65

SPO Abstracts

IS IT COST EFFECTIVE TO SCREEN TEENAGERS FOR GESTATIONAL DIABETES? A.Johnson: S.Kady Wilson: L.Hosmer: K.Franko-Filipasic,' L.Chambliss,Dept. of Db/Gyn, Maricopa Medical Center, & The Midwifery Service, Phoenix Memorial Hospital, Phx. AZ. OBJECI1VE: To determine the cost effectiveness of screening teenagers for gestational diabetes. STUDY DESIGN: We retrospectively reviewed the results of the prenatal testing for gestational diabetes in 350 women age 19 and younger who delivered at our institutions. The results of the 10 glucola and any 3° OTIs were recorded. Patients with an abnormal 3° G1T were identified and assigned a class of gestational diabetes as described by White. The cost of a I' glucola is $19.85 and for a 3' OTT is $52.00. The costs for testing this group of 350 patients was calculated. The cost to identify each patient with gestational diabetes was determined by dividing the number of cases identified by the total cost of testing. RESULTS: 9 of the 350 patients screened had a I' glucola > 140 mg% and had a 3' OTT. 3 of these patients met criteria for diabetes. All 3 were Class Al with normal fasling glucoses throughout their pregnancies on diet alone. The total cost of 350 one hour glucolas was $6,947.50. The cost of the 9 three hour OTTs was $468.00. CONCLUSION: To screen 350 pregnant teenagers for gestational diabetes it cost $7,204.50 and identified only 3 gestational diabetics. All were Class AI and none required insulin. Lucas et. aI.(I) reported that patients with Class

January 1994

Am J Obstet Gynecol

67

Al gestational diabetes did not have significant obstetrical or perinatal

morbidity. In 1991533,483 patients 19 years and younger delivered in the U.S.(2) The cost of testing for gestational diabetes for this group would be approximately II million d
Obstet Gynecol 1993;82:260-265 2. Center for Disease Control yearly reports: 1991

66

CHORlOAMNIOIIITlS: THE ASSOCIATIOII OF AIIIKlRMAL FHR PATTERNS AND NEONATAL ACIDEMIA_ Dax\ l. Gilstrap. ~~p~e~t>"}GY;: ~~~~ _S-o:o~n~ut~:..:.:'t":;~YM~: Ctr., Dallas, TX OBJECTIVE: I t has been previ ousl y reported that the majority of newborns with acute chorioarrnionitis do not have low Ua pHis «7.20). We sought to determine whether selected variables, specifically FHR abnormalities, in pregnancies coq>licated by chorioamionitis could predict neonatal acidemia (i .e., cord pH < 7.20)_ METHOOS: During a 6-month period, mothers diagnosed with chorioamnionitis were prospectively identified in labor. Following del ivery, fetal monitor strips and hospital course were reviewed for both mother and neonate_ The Cox Proportional Hazards ModeL weighs independent variables according to their duration of exposure. Independent variables included; bi rthwei ght, loss of vari abil lty, absence of fetaL heart rate accelerations, presence of fetal tachycardia, and absence of severe and late fetal heart rate decelerations. Duration of exposure for the statistical model was based upon diagnosis to del ivery. . RESULTS: 197 consecutive cases of chorioannionitis were prospectively diagnosed for an incidence of 2_1% in our population. Independent variables and their relative risks for neonatal acidemia are 1isted below. CI (95%) RB. f Birthweight «2500 gms) _5-3.3 1.3 .60 _3-2.5 loss of var i abi 1i ty .85 -17 .4-1_7 .84 .63 ~~~~?g~a~f o~~vere/late No accelerations _01 2.8 1_3-6.1 Mi ld tachycardia (160-180) _41 .8 .4-1.4 Mod/severe tachycardia (~181) _28 1.5 -1-3.2 The Lack of fetal heart rate acceLerations was the only variabLe significantly associated with neonatal acidemia (pH < 7.20). COIICLUSIOII: In pregnancies compl icated by chorioamnionitis, only the absence of fetal heart rate accelerations could predict neonatal acidemia_ To our knowledge, this is the only report to evaluate fetal heart rate patterns according to their length of exposure.

68

MAGNESIUM SULFATE VERSUS PHENYTOIN (PRY) FOR SEIZURE PREVENTION IN AMYGDALA KINDLED RATS. C.A. Standley', S.M_lrtenkauF, L. Stewart", B. Mason, D.B. Cotton. Dept of Ob/Gyn, Hutzel HospitallWayne State University, Detroit, MI. OBJECTIVES: MgSO, is widely used for seizure prophylaxis in preeclampsia-eclampsia. However, its anticonvulsant effects in other types of seizures has not been proven. In the preseut study, we compare the anticonvulsant effects of therapeutic blood levels of MgSO, and PHY in amygdala kindled rats. STUDY DESIGN: 18 male rats had a bipolar electrode stereotaxically implanted into the central nucleus of the amygdala. Following recovery, an afteroischarge (seizure) threshold was determined for each rat. Rats were stimulated daily at their seizure thresholds ( = kindling) until 3 consecutive generalized tonic-clonic seizures occurred. Kindled rats randomly received one of the following intravenous injections in a volume of 1.5 mllkg: saline, MgSO, (30, 60 or 90 mglkg) or PHY (12.5, 25 or 50 mglkg). Doses were based on preliminary data demonstrating blood levels of 16.2 mg/dl for 25 mglkg PHY and 4.2 mg/dl for 60 mglkg MgSO,. Fifteen minutes following irijectiou, rats were stimulated at their seizure thresholds and seizure activity was assessed. Statistical comparisons were done using ANOVA and post hoc comparisons where appropriate. RESULTS: Magnesium sulfate had no effect on any of the seizure parameters assessed. PHY significantly reduced seizure duration (p< .05), duration of postictal depression (p< .05) and behavioral seizure stage (p < .05). CONCLUSIONS: Amygdala kindled seizures are more potently inhibited by PHY than magnesium sulfate. This suggests that magnesium may not be a very effective treatmeut in situations where repetitive seizure activity has been experienced.

AMNIOTIC FLUID YOLIIIE flUCTUATIONS REflECT RECENT GLYCEMIC STATUS IN GESTATIONAL DIABETICS. I. Bar-Havax • S.A. Scarpelli x , Y. 8arnhardx , H. Y. Divon. Dept. Of Ob-Gyn, Albert Einstein College of Hedicine, Bronx, N.Y. OBJECTIVE: To determine the association between the sonographically derived amniotic fluid index (AFI) and recent gl ucose status in gestati onal di abetes meTT1tus (GDH). STWY DESIGN: The following variables were prospectively coTTected in 205 GOHs (who underwent 7 daily capillary glucose determinations with a memory glucometer) between 2/1/92 7/31/93: AFI, mean blood glucose 1 day prior and 1 week prior to the ultrasound examinations (HBGI-D and HBGI-W, respectively). and percent hyperglycemia (~120 mg%) 1 day prior and 1 week prior to the ultrasound examinations (PHI-D and PHI-W, respectively). ATT patients demonstrating at least 1 AFI measurement within the normal range (ie, Scm < AFI ~ 20cm) and at least 1 elevated measurement (ie, AFI > 20cm) formed the study population. With each patient serving as her own control. glucose indices preceding normal and elevated AFls (N-AFJ and E-AFI. respectively) were compared with the use of Student's t-test (n=39). In addition, AFls for each patient were plotted against gestational age to determine whether there was a consistent trend. RESULTS: AFI(cm) MBGI-D(mg%) M8GI-W(IIIg%) PHI-D(%) PHI-W(%) (Mean±SD) (Mean±SD) (Mean±SD) (Mean±SD) (Hean±SD) E-AFI 22±2 1I8±16 1I6±13 37±24 37±17 N-AFI 13±3 102±12 100±1I 17±18 18±14 p<0.0016 p