Abstract
The wide variation in cerebral and facial phenotypes and the recognized etiologic heterogeneity of holoprosencephaly (HPE) contribute to the observed inter‐study heterogeneity. High lethality during the early stages of embryonic and fetal development makes HPE detection age dependent. By reviewing 21 HPE epidemiologic articles, the observed prevalence rate differences can be largely explained by the pregnancy outcome status of the studied cohort: livebirth, stillbirth, and terminations of pregnancy (TOPs): lower than 1 per 10,000 when live and still births were included, higher when TOPs were included, and between 40 and 50 per 10,000 in two classical Japanese studies on aborted embryos. The increasing secular trend observed in some studies probably resulted from an increasing use of prenatal sonography. Ethnic variations in birth prevalence rates (BPRs) could occur in HPE, but the available data are not very convincing. Higher BPRs were generally observed in the less favored minorities (Blacks, Hispanics, Pakistanis), suggesting a bias caused by a lower prenatal detection rate of HPE, and consequently less TOPs. Severe ear defects, as well as microstomia, were part of the spectrum of HPE. Non‐craniofacial anomalies, more frequently associated with HPE than expected, were genital anomalies (24%), postaxial polydactyly (8%), vertebral defects (5%), limb reduction defects (4%), and transposition of great arteries (4%). The variable female predominance, found in different HPE studies, could also depend on the proportion of early conceptions in each study sample, as males are more likely to be lost through spontaneous abortions. © 2010 Wiley‐Liss, Inc.