590: Successfully overcoming barriers to genetic counseling through national telephone genetic counseling service

589 First-trimester prediction of preeclampsia using soluble P-selectin,follistatin-related protein 3, complement 3a, soluble TNF receptor type 1, PAPP-A, AFP, inhibin A, placental growth factor, uterine artery Doppler and maternal characteristics Lorraine Dugoff, Howard Cuckle, Nicholas Behrendt, Darleen Cioffi-Ragan, Leslie Myers, John Hobbins University of Pennsylvania, Obstetrics and Gynecology, Philadelphia, PA, Columbia University Medical Center, Obstetrics and Gynecology, New York, NY, University of Colorado, Obstetrics and Gynecology, Aurora, CO OBJECTIVE: To estimate the efficiency of first-trimester serum markers and uterine artery Doppler (UtA) in the prediction of severe preeclampsia (SPE). STUDY DESIGN: This is a prospective cohort study of patients presenting for aneuploidy screening between 11-14 weeks’ gestation. Maternal serum levels of soluble P-selectin, follistatin-related protein 3, complement 3a, soluble TNF receptor type 1, PAPP-A, AFP, inhibin A(inhA) and placental growth factor(PlGF) were measured by immunoassay and mean UtA resistance indices(RI) were calculated. The primary outcome was SPE. A secondary outcome of all PE including mild preeclampsia was also evaluated. Marker levels were expressed in multiples of the normal median (MOM), adjusting for gestation, weight, smoking and ethnicity. Logistic regression analysis was used to model performance. Maternal characteristics identified as significant in a univariate analysis were included in the models. Predicted sensitivity, specificity, and area under the receiver-operating characteristic curves were used to compare the screening efficiency of the models. RESULTS: Of 1028 patients, 26(2.5%) had SPE and 56(5.4%) had PE. Median PlGF levels were lower in patients with SPE (0.83 MoM vs 0.98 MoM, P 0.05) and PE(0.90 MoM vs 0.98 MoM, P 0.05) compared to those without PE. AFP was significantly higher (1.15 MoM vs 1.00 MoM, P 0.05; 1.20 MoM vs 0.99 MoM, P 0.01). Median RI was significantly increased in PE (1.05 MoM vs 1.00 MoM, P 0.05) but not in SPE; similarly for inhA (1.09 MoM vs 0.99 MoM, P 0.05). For a fixed false positive rate of 20%, PlGF and AFP identified 49% of patients with SPE; combining these with prior risk factors did not improve the model efficiency. The detection rate for PE was 44% modeling with PlGF, AFP, inhA and RI. CONCLUSION: First-trimester PlGF and AFP are each modestly predictive of SPE, but combining both improves detection.