Abstract

Maria Florea, *Dominic Gabriel Iliescu, Roxana Dragusin, Alex Stepan, Lucian Zorila, Stefania Tudorache

ABSTRACT

Objectives: The aim of this research is to evaluate the modeled predictive value of different features of the fetal anatomy in the detection of isolated open spina bifida, at first-trimester scan. Methods: We retrospectively assessed audited randomized two-dimensional and three-dimensional datasets from 6 cases of open spina bifida and 60 normal fetuses, with no structural or chromosomal associated anomalies, scanned at the same age (12 weeks +4 days). We analyzed direct and indirect signs, using both continuous and categorical variables. Effects were quantified using multiple logistic regression analysis to develop a mathematical model. The performance of screening was determined by Receiver Operating Characteristics (ROC) curve analysis. Results: We found many significant markers’ differences between the two groups. We determined the optimum set of features and their effect on OSB diagnosis. The intracranial subjective markers obtained in axial planes (the dry brain phenomenon, position of the Sylvius aqueduct and the crush sign) had the highest potential on improving OSB diagnostic accuracy. Conclusions: Application of intracranial markers performs better than the direct spine assessment. Application of cranial axial views based screening model (the lateral third-ventricle view and the transthalamic view, used for fetal head biometry) resulted in the highest predicted detection rates, thus a better screening performance than using sagittal views. Following a complete protocol in the FT might be the only efficient way to screen for isolated spina bifida at the 11–13-week scan. The findings should be confirmed in population-based prospective studies.