Head growth in patients with myelomeningocele treated with prenatal and postnatal surgery

Abstract

OBJECTIVE The need for permanent CSF diversion is lower in patients who have undergone prenatal surgery for myelomeningocele (MMC) than in those who have undergone postnatal closure. Differences in brain development and head growth between treatment groups are not known, particularly for those who do not require surgical treatment for hydrocephalus. The objective of this study was to determine differences in head growth and to generate MMC-specific head circumference (HC) growth curves for patients who underwent either prenatal or postnatal surgery. METHODS The authors retrospectively identified patients from St. Louis Children’s Hospital who were treated for MMC between 2016 and 2021. HC data were obtained from birth until the most recent follow-up or hydrocephalus treatment. Nonlinear least-squares regression analysis was performed to fit the data into four models: two-term power, Gompertz, West ontogenetic, and Weibull. Subsequently, the curves were assessed for their utility in predicting hydrocephalus treatment. RESULTS Sixty-one patients (29 females [47.5%], 25 [41%] underwent prenatal surgery, mean gestational age at birth 36.6 weeks) were included in the study. The Weibull model best fit the HC data (prenatal adjusted R2 = 0.95, postnatal adjusted R2 = 0.95), while the Gompertz model had the worst fit (prenatal adjusted R2 = 0.56, postnatal adjusted R2 = 0.39) across both cohorts. Prenatal MMC repair patients had significantly larger HC measurements than their postnatal repair counterparts. The 50th percentile of the Weibull curve was determined as a useful threshold for hydrocephalus treatment: children with HC measurements that crossed and remained above this threshold were significantly more likely to have hydrocephalus treatment regardless of time of MMC repair (prenatal relative risk [RR] 10.0 [95% CI 1.424–70.220], sensitivity 85.7% [95% CI 0.499–0.984], and specificity 82.4% [95% CI 0.600–0.948]; postnatal RR 4.750 [95% CI 1.341–16.822], sensitivity 90.5% [95% CI 0.728–0.980], and specificity 75.0% [95% CI 0.471–0.924]). The HC growth curves of the MMC patients treated prenatally were significantly larger than the WHO HC curves (p < 0.001). CONCLUSIONS The Weibull model was identified as the HC growth curve with the best fit for MMC patients and serves as a useful predictor of hydrocephalus treatment. For MMC patients with hydrocephalus, prenatal repair patients fit the model well but postnatal repair patients did not, potentially indicating different mechanisms of hydrocephalus development. Those treated prenatally had significantly larger HC measurements compared with both the general population and those treated postnatally. Further study is needed to understand the long-term cognitive outcomes and optimal management of clinically asymptomatic patients with large HC measurements who were treated prenatally for MMC.