evidence of brain-sparing persists after treatment of twin-reversed arterial perfusion sequence Jodie Votava-Smith, James Cnota, Susan Gomien, Mounira Habli, William Polzin, James Van Hook, Timothy Crombleholme, Erik Michelfelder Cincinnati Children’s Hospital Medical Center, Fetal Cardiology Program, The Heart Institute, Cincinnati, OH, Cincinnati Children’s Hospital Medical Center, Fetal Care Center, Division of Pediatric General and Thoracic Surgery, Cincinnati, OH, Good Samaritan Hospital, Obstetrics and Gynecology, Cincinnati, OH, University of Cincinnati, Obstetrics and Gynecology, Cincinnati, OH OBJECTIVE: Pump twins of twin-reversed arterial perfusion (TRAP) sequence may have altered cerebrovascular tone to preserve brain oxygenation in response to highly desaturated umbilical venous return from the acardiac twin. We sought to characterize the cerebroplacental Doppler patterns in pump twins before and after therapeutic radiofrequency ablation (RFA) of the acardiac twin. STUDY DESIGN: We retrospectively reviewed pre and post-RFA echocardiograms of TRAP fetuses performed between 2/04 and 2/11. Doppler evaluation included estimation of pump twin combined cardiac output indexed for weight (CCO), and pulsatility index (PI) of the middle cerebral artery (MCA) and umbilical artery (UA). PI zscores were calculated. Paired t-tests were used to compare pre and post-RFA Doppler data. RESULTS: 35 fetuses had complete pre-RFA Doppler datasets. Mean gestational age (GA) was 20.6 / 2.5 weeks. Mean z-score for UA-PI was 2.8 / 1.1, with 80% of fetuses having UA-PI z-score 2. Mean z-score for MCA-PI was 1.0 / 1.0, including 17% of fetuses with MCA-PI z-score 2 and none with MCA-PI z-score 2. Estimated fetal weight was 5th percentile for GA in 34/35 (97%). Post-RFA Doppler data was available in 12 pump twins, with follow-up study done a median of 2 days (interquartile range 1,5) post-RFA. Following intervention, pump twin CCO decreased but there was no significant change in MCA-PI, UA-PI, or CPR (Table 1). CONCLUSION: Prior to RFA therapy, the majority of the pump twins demonstrate low CPR secondary to abnormally high UA-PI rather than low MCA-PI. These findings persist in short term follow-up after RFA, despite a significant decline in CCO indicating interventional success. These results suggest elevated placental vascular resistance, which may be due to abnormal placental vascular connections which persist after cord ablation of the acardius, versus placental vascular remodeling secondary to chronic hypoxia. Further studies to characterize the placental pathology of this condition are warranted. 351 Accuracy of fetal weight estimation by ultrasound in periviable deliveries John Ethridge, Judette Louis, Brian Mercer MetroHealth Medical Center-Case Western Reserve University, Division of Maternal Fetal Medicine, Department of Obstetrics and Gynecology, Cleveland, OH OBJECTIVE: We sought to evaluate the accuracy of fetal weight estimation (EFW) by ultrasound (US) in women delivering near the limit of viability, and to determine factors that impact this accuracy. STUDY DESIGN: Retrospective chart review was performed for women delivering between 22 0/7 and 25 6/7 weeks gestation (GA). Individual patient records were reviewed to assess if an US was performed for EFW within 7 days of delivery. If so, additional potentially relevant clinical factors were abstracted (see results) and training level of person performing the ultrasound documented. These factors were examined to determine their impact on the accuracy of EFW. EFW and birth weights (BWT) were entered into an NICHD algorithm to predict survival, and were then compared between each other. RESULTS: 100 patients meeting inclusion criteria (2/2007 8/2011) were included. The cohort was 60% African-American, with median parity 1. Mean [S.D.] maternal age was 25.9 [6.9] yrs, and BMI 30.0 [7.4] kg/m2. Mean absolute percent difference (accuracy) between EFW and BWT was 10.2% (95% CI 8.0 12.4). Linear regression analysis revealed no correlation between EFW accuracy and GA, maternal age, or BMI. While significant associations were seen between BWT and accuracy (p 0.019; 1% decrease in absolute error per 100 gm increase in BWT), and in mean % error and BWT (p 0.001, 3% overestimation of BWT per 100 gm decrease in BWT), these were not clinically meaningful. ANOVA found no difference in accuracy based on race, fetal presentation, oligohydramnios, parity, smoking, prior CD, or level of training of the examiner. Mean predicted survival for the cohort from the NICHD algorithm was 43.1% using actual BWT compared with 43.6% using EFW (p 0.91). A difference greater than 20% in predicted survival was seen in only 2 cases. CONCLUSION: EFW in periviable deliveries is affected by BWT, but estimated survival based on ultrasound EFW within 7 days of delivery is similar to that based on actual BWT.