S188 SMFM Abstracts 665 GROUP B STREPTOCOCCUS INFECTION INDUCES TROPHOBLAST APOPTOSIS OZLEM EQUILS1, KELLY DORAN2, JAMES MCGREGOR3, CALVIN HOBEL4, 1University of California, Pediatrics, Los Angeles, California, 2University of California, San Diego, Pediatrics, San Diego, California, 3University of Southern California, Obstetrics and Gynecology, Los Angeles, California, 4University of California, Los Angeles, Los Angeles, California OBJECTIVE: Group B streptococci (GBS) is one of the leading causes of prematurity, and neonatal infection. Here, we assessed the cytotoxic effect of GBS infection on trophoblasts. STUDY DESIGN: We used JEG3 syncytiotrophoblast and 3A first trimester cytotrophoblast cell lines and either infected them with GBS or treated them with GBS extract. In order to assess whether trophoblast invasion was necessary for GBS induced trophoblast death, we infected JEG3 cells with invasiveness associated gene (iag) mutant GBS. In order to assess the effect of B-hemolysin on GBS induced torphoblast death, we treated JEG3 cells with GBS extract obtained from B-hemolysin mutant GBS. Trophoblast death was assessed by measuring supernatant lactate dehydrogenase release. Cell invasion was assessed by microscopic examination. RESULTS: We observed that in JEG3 syncytiotrophoblast and 3A first trimester cytotrophoblast cell lines both infection with GBS and treatment with GBS cell wall extract induced cell death as compared to uninfected and media treated cells respectively. We observed that iagGBS was poorly invasive in trophoblasts as compared to wild type bacteria; however iagGBS infection was significantly cytotoxic in JEG3 trophoblast cells. We observed that treatment of JEG3 trophoblast cells with cell extract from B-hemolysin deleted mutant was not cytotoxic. CONCLUSION: Trophoblasts provide a barrier between fetal circulation and placenta. Our data suggests that GBS invades and induces trophoblast cell death; which may impair fetal-placental barrier and consequently allow bacterial penetrance into fetal circulation. GBS-induced trophoblast death may provide an alternative mechanism for GBS-pathogenesis in neonatal GBS infection as well as preterm delivery.
667 THE ROLE OF FETAL VIRAL INFECTION IN THE DEVELOPMENT OF ADVERSE ALASTAIR MACLENNAN1, PREGNANCY OUTCOMES CATHERINE GIBSON1, PAUL GOLDWATER2, ERIC HAAN3, KEVIN PRIEST4, GUSTAAF DEKKER5, 1Adelaide University, Obstetrics and Gynaecology, Adelaide, South Australia, Australia, 2 Women’s and Children’s Hospital, Microbiology and Infectious Diseases, Adelaide, South Australia, Australia, 3Women’s and Children’s Hospital, Department of Genetic Medicine, Adelaide, South Australia, Australia, 4Department of Health, Epidemiology Branch, Adelaide, South Australia, Australia, 5 Adelaide University, Maternal Medicine, Adelaide, South Australia, Australia OBJECTIVE: To investigate the role of fetal viral infection in the development of a range of adverse pregnancy outcomes, including pregnancy-induced hypertensive disorders (PIHD), antepartum haemorrhage (APH), intrauterine growth restriction !10th percentile (IUGR) and preterm birth (PTB). STUDY DESIGN: The newborn screening cards of 717 cases and 609 controls were tested for viral RNA and DNA from enteroviruses and herpesviruses using polymerase chain reaction (PCR). The herpesviruses were detected using two PCRs, one detecting nucleic acids from HSV-1, HSV-2, EBV, CMV and HHV-8, hereafter designated Herpes group A viruses, and the other detecting nucleic acids from VZV, HHV-6 and HHV-7, hereafter designated Herpes group B viruses. RESULTS: Detection of cytomegalovirus (CMV) DNA was significantly associated with preterm birth !28 weeks gestation (OR 1.62, 95% CI 1.022.57). The risk of developing PIHD was increased in the presence of Herpes group B viruses (OR 3.57, 95% CI 1.10-11.70), CMV (OR 3.89, 95% CI 1.679.06), any herpesvirus (OR 5.70, 95% CI 1.85-17.57) and any virus (OR 5.17, 95% CI 1.68-15.94). Viral infections were not associated with IUGR or APH without hypertension; the combined presence of pre-existing hypertension and any herpesvirus was significantly associated with growth-restriction (OR 5.70, 95% CI 1.17-27.73). CONCLUSION: Exposure to in utero viral infection is significantly associated with very preterm birth !28 weeks gestation, and also with pregnancyinduced hypertensive disorders, but not with IUGR or APH.
666 PYELONEPHRITIS DURING PREGNANCY: A CAUSE FOR AN ACQUIRED DEFICIENCY OF PROTEIN Z JYH KAE NIEN1, DEBRA HOPPENSTEADT2, OFFER EREZ3, JIMMY ESPINOZA3, ELEAZAR SOTO3, JUAN PEDRO KUSANOVIC3, CHONG POOJA MITTAL3, JAWED FAREED2, JOAQUIN SANTOLAYA3, JAI KIM4, ROBERTO ROMERO1, 1Perinatology Research Branch, NICHD, NIH, DHHS, Bethesda, Maryland, 2Loyola University Medical Center, Department of Pathology, Maywood, Illinois, 3Wayne State University School of Medicine, Department of Obstetrics and Gynecology, Detroit, Michigan, 4Wayne State University School of Medicine, Department of Pathology, Detroit, Michigan OBJECTIVE: Pyelonephritis during pregnancy has a more severe course than in the non-pregnant state. This has been attributed to the susceptibility of pregnant women to the Schwartzman Reaction (DIC). An acquired protein Z deficiency has been reported when there is excessive thrombin activity. The aim of this study was to determine whether pyelonephritis results in changes in the maternal plasma protein Z concentrations. STUDY DESIGN: A cross sectional study was designed and protein Z plasma concentrations were compared between normal pregnant women (n=71) and pregnant women with pyelonephritis (n=43), matched for gestational age. Protein Z plasma concentrations were measured by ELISA. Non-parametric statistics were used for analysis. RESULTS: 1) There was no change in plasma protein Z concentrations with advancing gestational age in normal pregnant women; and 2) patients with pyelonephritis had a significantly lower median plasma concentration of protein Z than patients with a normal pregnancy (median 2.1 ml/ml, range 0.43.2 vs. median 2.4 ml/ml, range 1.1-3.4; p=0.029). CONCLUSION: Pyelonephritis during pregnancy is associated with lower concentrations of protein Z. The most likely explanation for this observation is enhanced thrombin activity, which is known to cause an acquired protein Z deficiency. This state can contribute to the severity of pyelonephritis during pregnancy (e.g., development of ARDS).
668 HIVNET 024: NEUTROPHILIC AND MONONUCLEAR PLACENTAL INFILTRATION, ANTIBIOTICS AND OUTCOME IN AN AFRICAN POPULATION ROBERT GOLDENBERG1, VICTOR MUDENDA2, ELIZABETH BROWN3, TAHA TAHA4, 5 1 EPHATA KAAYA , University of Alabama at Birmingham, Obstetrics/Gynecology, Birmingham, Alabama, 2University Teaching Hospital, Pathology, Lusaka, Zambia, 3Fred Hutchinson Cancer Research Center, Seattle, Washington, 4Johns Hopkins University, Baltimore, Maryland, 5Muhimbili University, Pathology, Dar es Salaam, Tanzania OBJECTIVE: Histologic chorioamnionitis (HCA), usually defined as neutrophilic infiltration (NI) of the chorioamnion (CA), is associated with preterm birth (PTB), lower birthweights (BWT) and earlier gestational ages (GA). Antibiotics have not consistently reduced PTB. Fewer studies have evaluated NI in other placental sites or mononuclear infiltration (MI) at any site and their relationship to antibiotic use or PTB. Our goal is to evaluate the relationship of NI and MI to pregnancy outcome. STUDY DESIGN: A randomized trial of antibiotics to reduce PTB and perinatal HIV transmission was conducted at 4 African sites with negative results. Both HIV-infected (n=2098) and uninfected (n=335) women were given metronidazole 250 mg and erythromycin 250 mg at 20-24 wks tid for 7 days and metronidazole 250 mg and ampicillin 500 mg q 4 hrs during labor, or identical placebos. At delivery, the CA, decidua, umbilical cord and villi were evaluated for NI and MI. RESULTS: NI was common in the decidua (18%) and CA (28%) but less so in the cord (7%) and villi (0.1%). MI was common in the decidua (43%) and CA (34%) but not in the cord (0.1%) or villi (0.2%). NI in the membranes (29 vs 23% p!0.05) was more common in HIV infected women as was MI (35 vs 29% p=0.07). Decidual MI was associated with fewer PTB (20.1 vs 24.3% p=0.03) and longer time from randomization to delivery (p=0.04). The relationship between NI at various sites and outcome are shown in the table ()p!0.05). Antibiotics did not decrease NI or MI at any site. Cord
Membranes
Decidua
Outcome
No NI
NI
No NI
NI
No NI
NI
BWT(g)(x) GA(x) PTB ! 32 wks (%) PTB ! 37 wks (%) BWT ! 1500 g (%) BWT ! 2500 g (%)
2999 39.0 2.5 21.7 1.3 11.6
2779* 38.2* 7.0* 32.4* 4.9* 23.2*
3005 39.0 2.4 22.0 1.2 10.9
2926* 38.9 4.0 23.7 2.5* 16.5*
3003 39.1 2.2 21.2 1.0 11.4
2898* 38.4* 6.1* 28.3* 4.7* 17.0*
CONCLUSION: NI, but not MI, in the cord, decidua, and CA is associated with PTB at various GA cutoffs and BWT !1500g and !2500g, as well as decreased mean BWT and GA. HIV-infected women are more likely to have NI, perhaps explaining part of the poorer outcomes in those women. Antibiotics do not decrease NI or MI.