Biometric Assessment of Pathological Fetal Corpus Callosum: Ultrasound Versus MRI

BACKGROUND: Ultrasound and fetal MRI are both used to evaluate suspected corpus callosum abnormalities, but their measurement agreement remains poorly defined. This study assessed agreement between ultrasound and MRI measurements of corpus callosum length and thickness, and their impact on prenatal diagnosis and management. METHODS: This retrospective single-center study included fetuses referred for suspected corpus callosum or septum pellucidum abnormalities who underwent fetal MRI (2019-2023). Agreement was assessed in the 10 fetuses with a measurable MRI using intraclass correlation coefficients (ICCs) with 95% confidence intervals and Bland-Altman analysis. The effect of the inter-examination interval was evaluated using linear regression, and length measurements were standardized using gestational age-adjusted z-scores and compared with the growth expected from a reference curve over the interval. RESULTS: Fourteen fetuses were included. Mean gestational age was 27.8 ± 3.6 weeks at ultrasound and 31.8 ± 2.4 weeks at MRI (mean interval 4.1 ± 2.4 weeks). Corpus callosum length was greater on MRI (37 vs. 30 mm; 10 paired observations), with poor agreement (consistency ICC 0.29, 95% CI -0.38 to 0.76; agreement ICC 0.18, 95% CI -0.18 to 0.64). After z-score standardization, the mean difference was 0.55 SD (95% CI -0.99 to 2.09) and no association with the inter-examination interval was detected, but the limits of agreement remained wide (-3.7 to 4.8 SD). The raw difference (6.1 mm) was close to the growth expected over the interval (5.6 mm), which is compatible with a gestational age effect, although a modality effect cannot be excluded. Thickness measurements showed poor reproducibility. MRI modified the prenatal diagnosis in 11/14 cases and management in 8/14 cases in this population referred for suspected callosal or septal abnormalities. CONCLUSION: Because ultrasound and MRI were performed about 4 weeks apart, a modality effect could not be separated from physiological growth, and agreement between the two techniques was not established. Most of the raw difference was compatible with the effect of gestational age, which should be accounted for when comparing callosal biometry across examinations. Prospective studies with contemporaneous examinations in larger samples are needed.