571 CIRCULATING TOTAL CELL-FREE DNA LEVELS ARE INCREASED IN PREGNANT WOMEN WHO SMOKE ADAM URATO (F) 1 , INGA PETER 1 , GERALYN MESSERLIAN 2 , JACOB CANICK 2 , ANDREA PULKKINEN 3 , GEORGE KNIGHT 3 , YOUNG JU JEONG 1 , SABRINA CRAIGO 1 , DIANA BIANCHI 1 , 1 Tufts University-New England Medical Center, Boston, Massachusetts, 2 Brown University, Providence, Rhode Island, 3 Foundation for Blood Research, Scarborough, Maine OBJECTIVE: Cell-free DNA is believed to be a marker of cellular apoptosis and necrosis. Increased levels have been found in patients with lupus, pulmonary embolus, trauma, stroke, myocardial infarction, cancer, and in certain complications of pregnancy. In this study we wished to determine if smoking affects maternal serum cell-free DNA levels during pregnancy. STUDY DESIGN: A cross-sectional study was performed using stored second-trimester quadruple serum screening samples obtained from pregnant smokers and pregnant nonsmokers. Smokers and nonsmokers were matched for gestational age and storage time. Smoking amount was determined by maternal report and cotinine level. Glyceraldehyde-3-phosphate dehydrogen- ase (GAPDH) and DYS1 levels were determined using real-time quantitative PCR amplification to measure total and fetal cell-free DNA, respectively. RESULTS: Results are based on 117 pregnancies (27 smokers and 90 nonsmokers) for GAPDH assessment and a subset of 100 pregnancies with male fetuses (25 smokers and 75 nonsmokers) for DYS1 assessment. Smoking was associated with elevated GAPDH levels when compared with nonsmokers (median: 97,662 genome equivalents (GE)/ml vs. 38,217 GE/ml; p=0.017). Cotinine levels were correlated with GAPDH levels (p!0.001). DYS1 levels were higher in smokers versus nonsmokers (median: 41.2 GE/ml vs 40.3 GE/ml), but the result was not statistically significant (p=0.41). DYS1 levels did not correlate with cotinine level. CONCLUSION: Pregnant smokers have higher levels of total cell-free DNA compared with pregnant nonsmokers; these levels correlate with serum cotinine values. Further studies are needed to determine the specific maternal tissue that is damaged by smoking and results in the release of increased circulating cell-free DNA in pregnant women who smoke. 572 PATIENT DECISIONS REGARDING PRENATAL ANEUPLOID FISH ANDREA WRAY 1 , HELAIN LANDY 1 , JEANNE MECK 1 , 1 Georgetown University Hospital, Obstetrics and Gynecology, Washington, District of Columbia OBJECTIVE: The objectives of this study were (1) to determine the threshold of risk that most benefits offering prenatal aneuploid FISH and (2) if the availability of FISH results affects patient decisions. STUDY DESIGN: This retrospective chart review evaluated 707 patients presenting for genetic counseling and diagnostic prenatal testing between 1/1/ 2002 and 12/31/2004. Studied parameters included gestational age, indication for testing, aneuploid risk, procedure performed, FISH (whether offered, requested, and/or performed), result turn-around time, karyotype results, decision after obtaining results, and the timing of that decision. Patients who were offered FISH were compared to those not offered FISH (student T-test). RESULTS: Thirty abnormalities were detected; 5 undetectable by FISH. Of those detected, 5 abnormalities were not of clinical significance (2 mosaic tetraploid results on CVS and 3 inherited balanced rearrangements) and were not used in further comparisons. Seventeen patients elected pregnancy inter- ruption, 13 of whom had FISH performed. Of these 13, 7 ended the pregnancy following FISH alone and 6 following final G-banded karyotype. There was no difference between gestational age (P=0.65), risk (P=0.13), maternal age (P=0.69), gravidity (P=0.25) or parity (P=0.54) in the group who elected interruption following FISH results alone compared to the group who elected interruption following the final results. Using a risk cut-off of O1%, 9 of the 23 (83%) FISH-detectable anomalies would have been detected, with 317 FISH analyses performed. If a O0.5%, O2%, or O3% risk were used, the detection rates would have been 100%, 52%, and 35%, respectively, with 663, 118, or 66 FISH analyses, respectively, performed. CONCLUSION: Half of the patients elected to act on FISH results alone, allowing for a significantly shorter time between test and pregnancy interrup- tion (2 days vs 8 days). Using a risk cut-off of O1%, appears to optimize the detection rate and the yield of abnormal results (6% of total FISH analyses). 573 GENE EXPRESSION OF TRISOMY 21 PLACENTA USING EXPRESSION MICROARRAYS AND QUANTITATIVE RT-PCR ANAT JONISH 1 , ANAT BAR-SHIRA 2 , YIFAT OCHSHORN 3 , URI ROZOVZKI 2 , MIRIAM GOLDSTEIN 2 , YUVAL YARON 2 , 1 Tel Aviv University, Genetics, Tel Aviv, Israel, 2 Sourasky Medical Center, Prena- tal Diagnosis Unit & Genetic Institute, Tel Aviv, Israel, 3 Sourasky Medical Center, Lis Maternity Hospital, Dept Ob/Gyn, Tel Aviv, Israel OBJECTIVE: To identify genes that are differentially-expressed in trisomy 21-affected placentas compared to placentas of chromosomally normal fetuses. STUDY DESIGN: The study included CVS samples from 4 trisomy 21- affected pregnancies and 4 normal controls. Total RNA was extracted from cultured trophoblast and hybridized to AffymetrixÒ Human Genome U95Av2 Microarray containing w12,000 transcripts. For each transcript, a signal value and a detection call were generated using MAS5 and analyzed using Gene- SpringÒ software. Validation of selected genes was performed by quantitative RT- PCR. RESULTS: 5334 genes present in at least 4 out of 8 samples were chosen for non-supervised hierarchical clustering. This set of genes successfully discrim- inated between all trisomic and normal samples. ANOVA, using False Discovery Rate (FDR) for multiple comparisons correction, identified 65 genes that demonstrated a significant differential-expression between trisomic and normal samples. Two genes significantly over-expressed in trisomy 21 (MEST and LOX), and one under-expressed gene (MAT2A) were further studied by quantitative RT-PCR. This analysis confirmed over-expression of MEST (13.6-fold, p=0.036) and LOX (3.7-fold. p=0.025) (shown below), but did not validate under-expression of MAT2A. CONCLUSION: These data demonstrate that gene expression in trisomy 21- affected placentas significantly differs from chromosomally normal placentas. Further study is needed to assess the role of MEST and LOX in the pathophysiology of Down syndrome. This approach may facilitate the discovery of additional maternal serum markers of trisomy 21. SMFM Abstracts S163