Objective To examine the association of placental and fetal DNA copy number variants (CNVs) with fetal structural malformations (FSMs) in stillborn fetuses. Design A secondary analysis of stillbirth cases in the Stillbirth Collaborative Research Network (SCRN) study. Setting Multicenter, 59 hospitals in 5 geographic regions in the USA. Population 384 stillbirth cases of the SCRN study (2006-2008). Methods FSMs were grouped by anatomic system and specific malformation type (e.g., central nervous system, thoracic, cardiac, gastrointestinal, skeletal, umbilical cord and craniofacial defects). Single-nucleotide polymorphism array detected CNVs of at least 500kb. CNVs were classified into two groups: normal, defined as no CNVs>500kb or benign CNVs, and abnormal, defined as pathogenic or variants of unknown clinical significance. Main outcome measures The proportions of abnormal CNVs and normal CNVs were compared between stillbirth cases with and without FSMs using the Wald Chi-squared test. Results The proportion of stillbirth cases with any FSMs was higher among those with abnormal CNVs compared with those with normal CNVs (46.7% vs. 19.6%; p-value<0.001). The most common organ system-specific FSMs associated with abnormal CNVs were cardiac defects, followed by craniofacial and skeletal defects. A pathogenic deletion of 1q21.1 involving 46 genes (e.g., CHD1L) and a duplication of 21q22.13 involving 4 genes (SIM2, CLDN14, CHAF1B, HLCS) were associated with a skeletal and cardiac defect, respectively. Conclusion Specific CNVs involving several genes were associated with FSMs in stillborn fetuses. The findings warrant further investigation and may inform counseling and care surrounding pregnancies affected by FSMs at risk for stillbirth.

Objective To examine the association of DNA copy number variants (CNVs) with pathologic placental lesions (PPLs) in stillborn fetuses. Design A secondary analysis of stillbirth cases in the Stillbirth Collaborative Research Network case-control study. Setting Multicenter, 59 hospitals in 5 geographic regions in the USA. Population 387 stillbirth cases (2006-2008). Methods Using standard definitions, PPLs were categorized by type including maternal and fetal vascular, inflammatory and immune/idiopathic lesions. Using single-nucleotide polymorphism array, CNVs of at least 500 kb were detected. CNVs were classified into two groups: normal, defined as no CNVs>500 kb or benign CNVs, and abnormal, defined as pathogenic or variants of unknown clinical significance. Main outcome measures The proportions of abnormal CNVs and normal CNVs were compared between stillbirth cases with and without PPLs using the Wald Chi-squared test. Results Of 387 stillborn fetuses, 327 (84.5%) had maternal vascular PPLs and 60 (15.6%) had abnormal CNVs. Maternal vascular PPLs were more common in stillborn fetuses with abnormal CNVs compared with those with normal CNVs (81.7% vs. 64.2%; p=0.008). The proportions of fetal vascular, maternal/fetal inflammatory, and immune/idiopathic PPLs were similar among stillborn fetuses with abnormal CNVs compared to those with normal CNVs. Pathogenic CNVs in stillborn fetuses with maternal vascular PPLs spanned several genes with known relevant mechanisms. Conclusions Abnormal placental/fetal CNVs were associated with maternal vascular PPLs in stillborn fetuses. Findings may provide insight on the mechanisms of specific genetic abnormalities associated with placental dysfunction and stillbirth.

Preterm neonatal survival: what is the role of prognostic models?Elizabeth M McClure, PhD1Robert L Goldenberg, MD21Social, Statistical and Environmental Sciences, RTI International, Durham, NC2Columbia University, New York, NYEven before the 1960’s and the introduction of the specialty of neonatology, and continuing to the present, numerous efforts have been made to understand the relationship between newborn birthweight and the risk of mortality. (1) With the development of neonatal intensive care units (NICUs), attention to survival rates and neurologic outcomes among those at the lowest birthweights and gestational ages (GA) has grown. (2) Defining the lower limits of GA or birthweight associated with the neonatal outcomes is important for clinicians, families, and others to inform appropriate decision-making and clinical care.To predict newborn survival, numerous models have been developed to estimate risk at specific birthweights and/or GAs. To date, more than 35 have been published, almost exclusively from high-income countries with advanced NICU care. In a study published recently, van Beek et al sought to validate one of these predictive models from the United Kingdom (UK), deemed to be among the highest quality, with the objective of assessing its value for clinical use. (3)Van Beek et al used an independent Dutch population to validate survival among very preterm infants using the UK model’s parameters. Because they found relatively good performance, the authors’ concluded that the model could inform daily clinical practice. However, the generalizability of their results, especially to other populations differing by ethnicity or socioeconomic status, is questionable. The parameters for the model quality focused on birthweight, GA, and gender, but many other metrics (including the racial diversity, quality of care, etc.) were limited. In particular, the interventions available and utilized for obstetric and neonatal care were not specified, which would be important for their goal of clinical use of the model. Importantly, the quality of obstetric care is not considered. (4) Both the availability and quality of specific obstetric and neonatal interventions in any given setting may be among the most important factors impacting survival.Especially important for clinical considerations, long-term outcomes, including severe disabilities, were not addressed. Concerns about neurodevelopmental outcomes in infants at the lower limits of birthweight and GA are as or more important to parents and caregivers than survival. (5) It is thus unclear how this – or virtually any other model - can be useful for “daily clinical practice”.A better strategy to inform clinical care is for individual health-care facilities to maintain neonatal survival and neurological outcome statistics. These types of data within a specific context may be more helpful to physicians, including obstetricians and neonatologists, who often, together with parents and caregivers, make decisions related to interventions prior to delivery or during NICU care. Newborn outcomes, especially at the extreme lower limits of birthweight and GA, remains an area of intense interest. While models may provide some supportive information, it is difficult to imagine that these will ever replace clinical decisions informed by actual outcome data from the specific facility.Conflicts of interest: The authors declare no conflicts of interest.References:1. Goldenberg RL, Nelson KG, Dyer RL, Wayne JB. The variability of viability: the effect of physicians’ perceptions of viability on the survival of very low birth weight infants. Am J Obstet Gynecol 1982; 143:678-84.2. Bottoms SF, Paul RH, Mercer BM, MacPherson CA, Caritis SN, Moawad AW. Obstetric determinants of neonatal survival: antenatal predictors of neonatal survival and morbidity in extremely low birth weight infants. Am J Obstet Gynecol. 1999 80(3 Pt 1):665-9.3. Van Beek P.E, Groenendaal F, Onland W, Koole S, Dijk PH, Dijkmanet KP et al. Prognostic model for predicting survival in very preterm infants: an external validation study. BJOG (in press)4. Goepfert AR, Goldenberg RL, Hauth JC, Bottoms SF, Iams JD, Mercer BM Obstetrical determinants of neonatal neurological morbidity in < or = 1000-gram infants. Am J Perinatol. 1999;16(1):33-42.5. Iams JD, Mercer BM. National Institute of Child Health and Human Development Maternal-Fetal Medicine Units Network. What we have learned about antenatal prediction of neonatal morbidity and mortality. Semin Perinatol 2003:247-52.