Surveillance of multiple congenital anomalies; searching for new associations

Identifying new teratogens is crucial in preventing birth defects. Many teratogens cause a spectrum of anomalies rather than a single defect, so identifying frequently co-occurring anomalies can improve teratogen detection. Previous studies have shown that multiple congenital anomalies occur in about 16 per 10,000 births. This study aimed to leverage the large dataset from the EUROCAT network (covering 6,599,765 births from 32 registries between 2008 and 2016) to identify pairs or triplets of anomalies that occur more often than chance would predict. A developed algorithm aimed at automatically identifying these anomalies was used. The study also aimed to determine whether identified pairs/triplets were already known associations, or if they represented potential new associations requiring further research.

The paper references prior work on monitoring for multiple malformations in birth defect detection (Khoury et al., 1987, 1993, 1994; Kallen et al., 1999; Kalter, 1998), including the development and application of methods for analyzing registry data on infants with multiple malformations. Specific mention is made of the EUROCAT multiple congenital anomaly algorithm and its prior applications (Garne et al., 2011; Calzolari et al., 2014), along with studies on omphalocele and gastroschisis (Calzolari, 1995; Forrester et al., 2008), and the VACTERL association (Botto et al., 1997; van de Putte et al., 2020). The authors also draw upon research regarding the OEIS complex (Arteaga-Vázquez et al., 2019; Keppler-Noreuil, 2001) and the co-occurring defect analysis approach (Benjamin et al., 2019; Howley et al., 2023; Khoury et al., 1990).

The study utilized data from 32 EUROCAT congenital anomaly registries, encompassing 6,599,765 births from 2008 to 2016. After excluding cases with chromosomal or genetic syndromes, 123,566 cases with one or more major congenital anomalies remained for analysis. The EUROCAT multiple congenital anomaly algorithm, refined since 2004 by geneticists and pediatricians, classified cases into four groups: chromosomal syndromes, genetic and environmental syndromes, isolated anomalies, and multiple congenital anomalies. Approximately 10% of cases initially classified as potential multiples underwent review by three EUROCAT geneticists to ensure accurate classification. For each pair of anomalies, the odds ratio was calculated using a two-sided Fisher’s exact test, and the Benjamini-Hochberg procedure adjusted p-values to control the false discovery rate. Pairs with adjusted p-values < 0.05 were considered significant. A similar analysis was performed for all anomaly cases (not just multiples), using a more stringent p-value cutoff (< 0.01). Logistic regression models examined associations between three anomalies. The analysis was also conducted separately for males and females.

The EUROCAT algorithm identified 8804 cases (7.1%) with two or more major congenital anomalies in different organ systems. Analysis of cases with multiple congenital anomalies revealed 31 statistically significant positive associations between pairs of anomalies (adjusted p < 0.05). Of these, 20 were known associations (e.g., limb body wall complex, OEIS complex, VACTERL association). Eleven were considered potential new associations. Following a detailed literature review and examination of individual cases, six pairs remained classified as new associations: 1) encephalocele and Ebstein's anomaly; 2) microcephaly and cleft lip; 3) anophthalmos/micropthalmos and congenital cataract; 4) anophthalmos/micropthalmos and cleft lip; 5) anencephalus and gastroschisis; 6) hydrocephaly and hypoplastic right heart. No statistically significant associations were found among triplets of anomalies.

The identification of six potential new associations highlights the value of systematically searching for co-occurring congenital anomalies using large, high-quality datasets like EUROCAT. These findings warrant further investigation to explore potential underlying etiologies. The study's approach of analyzing multiple anomaly cases separately and then comparing with all anomaly cases offered a robust way to identify significant associations. The similar results between the two analyses reinforced the findings. However, the limitation of missing genetic information for some cases should be considered in the interpretation. The study contributes valuable insights into the complex interplay of congenital anomalies, potentially revealing novel teratogens or genetic mechanisms.

This study, using the extensive EUROCAT database, identified six novel associations of congenital anomalies that require further investigation. The study highlights the value of large-scale congenital anomaly surveillance in identifying new clusters of anomalies that may point to undiscovered teratogens or genetic mechanisms. Future research should focus on genetic and environmental risk factor analyses for these newly identified associations, potentially involving larger studies with access to more complete genetic data.

A limitation of the study was the lack of complete genetic information for all cases, hindering a more comprehensive understanding of the underlying mechanisms for some associations. The reliance on existing registry data and the inability to directly contact individuals for additional information may have limited the ability to fully characterize some cases. Also, the definition and classification of certain anomalies may vary across registries, potentially affecting the interpretation of results. This, however, was minimized via rigorous data cleaning processes and the standardized methodologies adopted by EUROCAT.