O-308 Characterizing the association between in vitro fertilization and VACTERL: a population-based assessment in 1.3 million live births

Is conception using in vitro fertilization (IVF) associated with the multiple congenital anomaly pattern referred to as VACTERL?

Children conceived through IVF were more than twice as likely to develop VACTERL compared to those naturally conceived (OR 2.57, 95% CI 1.93-3.39).

VACTERL (vertebral defects, anal atresia, cardiac defects, tracheo-esophageal fistula, renal anomalies, and limb abnormalities) is a multiple congenital anomaly pattern with a likely multifactorial origin, as few genetic or non-genetic risk factors have been identified. Some studies have suggested a potential link between IVF and VACTERL, yet this relationship has not been thoroughly investigated while accounting for other maternal characteristics. Given the rarity of VACTERL, large datasets with comprehensive information on fertility treatment parameters, maternal characteristics, and birth outcomes are essential to better understand this association.

We performed a population-based study of live births in four U.S. states (Massachusetts, New York, North Carolina, and Texas) between 2004 and 2018, linking birth records and birth defect data to IVF cycles reported in the Society for Assisted Reproductive Technology Clinic Outcome Reporting System (SART CORS). Naturally conceived births were selected as controls at a 10:1 ratio during the same period as the IVF birth. The study included a total of 1,315,867 live births.

All study children were linked to state birth defect registries, with defects classified using British Pediatric Association codes. VACTERL cases were identified by code 759.890 or the presence of at least three associated anomalies. Children with known chromosomal or other genetic syndromes were excluded. Logistic regression was used to assess the association between IVF conception and VACTERL. Covariates included maternal age, race/ethnicity, education, and state.

We identified 363 cases of VACTERL - 78 conceived through IVF (0.053%) and 285 naturally conceived (0.024%). Of these, 232 cases were identified using the diagnostic code 759.890, while 131 additional cases were identified by at least three of the six VACTERL associated anomalies. The most frequent VACTERL congenital anomalies were cardiac defects (54%), followed by renal anomalies (50%) and anal atresia (32%). In unadjusted analysis, IVF was significantly associated with an increased prevalence of VACTERL (OR 2.27, 95% CI 1.75-2.89). After adjusting for maternal socio-demographics, the association was further strengthened (OR 2.57, 95% CI 1.93-3.39) and supported in sensitivity analyses restricted to 1,226,080 singletons (OR 2.74, 95% CI 1.92-3.83). Evaluation of specific IVF treatment parameters (n = 141,882) revealed the highest prevalence of VACTERL in 58 of 102,099 births from fresh embryos (0.057% vs. 0.050% from thawed embryos), 41 of 69,765 births from ICSI (0.059% vs. 0.052% without ICSI), and 27 of 48,629 births from male infertility (0.056% vs. 0.055% without male infertility). Refining our multivariable model of IVF, we found that use of ICSI without a diagnosis of male infertility had the strongest association with VACTERL of all IVF groups (OR 3.04, 95% CI 1.90-4.65) compared to natural conceptions.

We were unable to assess familial clustering and terminations in the prevalence of VACTERL. However, our analyses are representative of live births and we accounted for maternal characteristics associated with both birth defect risks and the use of fertility treatment.

The use of IVF is associated with VACTERL in offspring, but the absolute risk for VACTERL in IVF-conceived offspring remains low. The mechanisms underlying this association are unclear, but altered DNA methylation patterns have been implicated in both IVF and VACTERL etiology.

No