development of a nomogram to predict preeclampsia

31 SOLUBLE ENDOGLIN, A NOVEL CIRCULATING ANTI-ANGIOGENIC FACTOR INPREECLAMPSIA (PE) RICHARD LEVINE1, CHUN LAM2, CONG QIAN3, KAI YU4,SHARON MAYNARD5, BAHA SIBAI6, ROBERTO ROMERO7, FRANKLIN EPSTEIN2,ANANTH KARUMANCHI8, 1NICHD, NIH, DHHS, Epidemiology Branch, Be-thesda, Maryland, 2Beth Israel Deaconess Medical Center, Medicine, Boston,Massachusetts, 3Allied Technology Group, Rockville, Maryland, 4NICHD,NIH,DHHS, Statistics Branch, Bethesda,Maryland, 5GeorgeWashingtonUni-versity Medical Center, Medicine, Washington, District of Columbia, 6Univer-sity of Cincinnati College of Medicine, Obstetrics and Gynecology, Cincinnati,Ohio, 7NICHD, NIH, DHHS, Perinatology Research Branch, Detroit, Michi-gan, 8Beth Israel Deaconess Medical Center, Medicine and Obstetrics, Boston,Massachusetts

OBJECTIVE: We recently noted that endoglin (E), a cell surface receptor forthe pro-angiogenic protein TGF-b, is upregulated in preeclamptic placentas.Therefore, we postulated that in preeclampsia excess soluble E (sE) is releasedfrom the placenta into the circulation through shedding of the extracellulardomain; sE may then synergize with sFlt1, an anti-angiogenic factor whichbinds placental growth factor (PlGF) and VEGF, to cause endothelialdysfunction. To test this hypothesis, we compared serum concentrations ofsE, sFlt1, and free PlGF throughout pregnancy in women who developedpreeclampsia to those of women with AGA/LGA infants who remainednormotensive during pregnancy (controls).

STUDY DESIGN: Nested case control study within the CPEP trial cohort ofhealthy nulliparas. We analyzed 861 serum specimens obtained throughoutpregnancy, but before labor, from 120 controls and the following preeclampsiacases: 120 with term PE, 72 with preterm PE, 9 with HELLP syndrome, 8 witheclampsia. The t-test was performed after log transformation.

RESULTS: Compared with GA-matched control specimens, sE and sFlt1were increased and free PlGF was decreased beginning 9-11 wks beforepreterm PE, reaching levels 5-fold (46.4 vs 9.8 ng/ml, P!.0001) and 3-foldhigher (6356 vs 2316 pg/ml, P!.0001) and 4-fold lower (144 vs 546 pg/ml,P!.0001), respectively, after PE onset. For term PE, sE was increasedbeginning 12-14 wks, free PlGF decreased beginning 9-11 wks, and sFlt1increased !5 wks before PE onset. Alterations in angiogenic factors weremore pronounced with early PE onset, PEC SGA, HELLP, or eclampsia. Theadjusted odds ratio for subsequent preterm PE was 9.8 (95% CI 4.5-21.5) forspecimens obtained at 21-32 wks which were in the highest quartile of controlsE concentrations (O7.2 ng/ml), as compared to all other quartiles.

CONCLUSION: Preeclampsia is associated with increased soluble endoglinand sFlt1 and decreased free PlGF months before onset of clinical disease.These findings suggest the importance of a circulating anti-angiogenic state inthe pathogenesis of preeclampsia.

33 THE ROLE OF FETAL INHERITED THROMBOPHILIA IN THE DEVELOPMENT OFADVERSE PREGNANCY OUTCOMES CATHERINE GIBSON1, NARD JANSSEN2,WILLEM KIST2, ALASTAIR MACLENNAN1, BILL HAGUE3, ERIC HAAN4, PAULGOLDWATER5, KEVIN PRIEST6, GUSTAAF DEKKER7, 1Adelaide University, Obstet-rics and Gynaecology, Adelaide, South Australia, Australia, 2VU UniversityMedical Center, Obstetrics and Gynaecology, Amsterdam, Netherlands,3Women’s and Children’s Hospital, Adelaide, North Adelaide, South Australia,Australia, 4Women’s andChildren’sHospital,DepartmentofGeneticMedicine,Adelaide, South Australia, Australia, 5Women’s and Children’s Hospital, Mi-crobiology and Infectious Diseases, Adelaide, South Australia, Australia, 6De-partment of Health, Epidemiology Branch, Adelaide, South Australia,Australia, 7AdelaideUniversity,MaternalMedicine, Adelaide, SouthAustralia,Australia

OBJECTIVE: To investigate the role of fetal inherited thrombophilia in thedevelopment of a range of adverse pregnancy outcomes, including pregnancy-induced hypertensive disorders (PIHD), antepartum haemorrhage (APH), in-trauterine growth restriction!10th percentile (IUGR) andpretermbirth (PTB).

STUDY DESIGN: 717 cases and 609 controls were genotyped for Factor VLeiden (FVL, G1691A), Prothrombin gene mutation (PGM, G20210A), andMethylenetetrahydrofolate reductase (MTHFR) C677T and MTHFR A1298Cusing genomic DNA extracted from newborn screening cards.

RESULTS: For babies born!28 weeks gestation, PGM was associated withan increased risk of IUGR(OR6.40, 1.66-24.71) andAPHwith IUGR(OR6.35,1.63-24.75). Homozygous PGM also increased the risk of PIHD with IUGRfor term-born babies (OR 50.81, 1.75-1476.90). Homozygosity for MTHFRA1298C was associated with an increased risk of IUGR for babies born 28-31weeks gestation (OR 4.00, 1.04-15.37), and with APH and IUGR for babiesborn !32 weeks gestation (OR 3.57, 1.09-11.66). MTHFR C677T was associ-ated with a reduced risk of PTB and IUGR (OR 0.52, 0.28-0.96) for babies born32-36 weeks gestation. Homozygous FVL was associated with an increasedrisk of PIHD with IUGR for term-born babies (OR 37.15, 1.33-1041.30), butdecreased the risk of PTB!32 weeks gestation (OR 0.55, 0.31-0.98). There wereno associations with any thrombophilic polymorphism and APH alone.

CONCLUSION: These results suggest that some fetal thrombophilic poly-morphisms may be related to adverse pregnancy outcomes, in particularIUGR, but this may not be the only association. Further studies matchingmaternal and fetal genotypes are required to investigate if both are needed forthe adverse pregnancy outcome phenotype to be expressed.

34 BONE DENSITY CHANGES IN WOMEN RECEIVING THROMBOPROPHYLAXIS IN PREG-NANCY HOLLY CASELE1, ELAINEHANEY2, ANDRA JAMES3, KAREN ROSENE-MONTELLA4,MICHAEL CARSON5, 1Northwestern University, Maternal Fetal Medicine, Evan-ston, Illinois, 2EvanstonNorthwesternHealthcare,Evanston, Illinois, 3DukeUni-versity, Obstetrics and Gynecology, Durham, North Carolina, 4BrownUniversity, Providence, Rhode Island, 5St. Peters University Hospital, InternalMedicine, New Brunswick, New Jersey

OBJECTIVE: To compare the effects on bone mineral density (BMD) ofunfractionated heparin (UH) and Enoxaparin sodium (LMWH) in patientsrequiring thromboprophylaxis in pregnancy.

STUDY DESIGN: Pregnant patients requiring thromboprophylaxis wererecruited in this multi-center, prospective randomized controlled trial andafter informed consent, were randomly assigned to receive either UH orLMWH. Bone mineral density was measured at the proximal femur uponenrollment or at the beginning of the second trimester and again shortly afterdelivery.

RESULTS: One hundred and seventeen women were enrolled from 9 centers.Twenty two patients discontinued study participation early. Eleven patientsexperienced a spontaneous loss, the remainder were noncompliant, self-withdrew or were switched to therapeutic anticoagulation. Six patients didnot complete the second BMD measurement. The UH and LMWH groupswere similar with regard to age, parity, racial distribution and treatmentduration.There was no difference in the change in BMD at the femoral neck(p=.113) or total proximal femur (p=.917) between groups. Only 2.4% (1/41)patients receiving UH and 4.2% (2/48) receiving LMWH (p=1.0) experiencedbone loss O10% at the proximal femur.

CONCLUSION: The incidence of clinically significant bone loss (O10%) inthe proximal femur in women receiving thromboprophylaxis in pregnancy isless than previous clinical estimates and appears to be similar regardless ofwhether the patient receives LMWH or UH.

S14 SMFM Abstracts

32 DEVELOPMENT OF A NOMOGRAM TO PREDICT PREECLAMPSIA ROMAN ROUZIER1,GILLES KAYEM1, STEPHANIE DEIS1, CHRISTOPHE MASSON1, BASSAM HADDAD1,1University Paris 12 - Creteil, Obstetrics and Gynecology, creteil, France

OBJECTIVE: To create a nomogram for the individual prediction ofpreeclampsia (PE).

STUDY DESIGN: In a prospective population-based study that included 4777patients, PE occurred in 2.4%. Age, body mass index, parity, previouspreeclampsia (PPE), chronic hypertension, diastolic blood pressure (DBP)and proteinuria at first visit, and second trimester umbilical artery doppler(UAD) resistance index (RI) data were used to develop and calibrate anomogram based on multivariable logistic regression model.

RESULTS: Based on multivariable analysis, nulliparity (P!.002), PPE(P!.004), DBP (P!.0001) and UADRI (P=.06) were introduced into anomogram (figure). Based on these variables, the nomogram had gooddiscrimination (area under the ROC curve = .73, P!.01) and calibration(Unreliability index = �5.2!10-4). This nomogram was validated by boot-strapping.

CONCLUSION: Our nomogram predicts probability of preeclampsia.