Data-driven personalized surgical treatment selection for metopic and sagittal craniosynostosis

Although the choice of surgical treatment for craniosynostosis is based on patient age and phenotypic severity, there are no tools to objectively identify optimal treatment approaches. The authors present a quantitative study of how age, sex, and phenotypic severity affect the surgical outcomes of patients with midline craniosynostosis to support treatment selection.

The authors collected pre- and postsurgical 3D photograms of patients with nonsyndromic single sagittal (n = 280) or metopic (n = 69) craniosynostosis, which they grouped into endoscopic strip craniectomy (ESC) or cranial vault remodeling (CVR) cohorts. The authors compared their surgical outcomes using the Head Shape Anomaly (HSA) index, Craniosynostosis Risk Score (CRS), and a novel Craniosynostosis Severity Index (CSI) that provides improved sensitivity to local malformations. Then, the authors quantitatively modeled the effect of age, sex, and presurgical severity on treatment outcomes and used their findings to create a decision tool to support treatment selection.

ESC and CVR improved all morphology metrics in patients with sagittal craniosynostosis (p < 0.001). In metopic craniosynostosis, ESC improved the CRS (p < 0.001) and CSI (p = 0.01), CVR improved the CRS (p = < 0.001) but not the CSI (p = 0.15), and neither ESC (p = 0.10) nor CVR (p = 0.83) improved the HSA index. The authors' model showed that ESC outperforms CVR in metopic and sagittal craniosynostosis before 3 and 5 months of age, respectively, as well as in nonsevere phenotypes of sagittal craniosynostosis before 8 months of age.

The authors' quantitative results can support personalized surgical treatment selections for patients with midline craniosynostosis and constitute a needed step to reduce the variability in treatment approaches and outcomes.