Fourth ventricular roof angle does not predict surgical outcome in paediatric patients with Chiari I malformation.

Abstract

Purpose: Pre-operative fourth ventricle roof angle (FVRA) has recently been highlighted as a novel radiographic predictor of clinical severity in Chiari type-I malformation (CM-I) as reported by Seaman et al. (J Neurosurg Pediatr 25:1-8, 2021). This has led to suggestions that FVRA could be included in algorithms to determine indications for surgery. We aimed to test the accuracy of FVRA as a predictor of clinical severity and its effect on post-operative outcome in a large retrospective cohort of paediatric patients who underwent decompression for CM-I.

Methodology: Patients undergoing craniocervical decompression for CM-I at the Queensland Children's Hospital and Mater Hospital, Brisbane, between 2006 and 2018 were included. Data was collected from 66 patients aged 0-18 years. Post-operative outcome was assessed by calculating Chicago Chiari Outcomes Score (CCOS) at follow-up. The FVRA was taken from a mid-sagittal T1-weighted MRI as previously described (Seaman et al. in J Neurosurg Pediatr 25:1-8, 2021). The angle is subtended between superior and inferior medullary velum.

Results: Whilst results from Seaman et al. demonstrated a strong correlation between a FVRA > 65° and symptomatic CM-I, this did not translate to our dataset (p = 0.61). Additionally, pre-operative FVRA is not useful as a predictive tool for post-surgical outcome as assessed by CCOS (p = 0.50), and post-operative reduction in FVRA did not correlate with improved outcomes (p = 0.81). We did note significantly worse outcomes in patients presenting with pre-operative brainstem dysfunction (p = 0.03).

Conclusion: Paediatric CM-I is a challenging cohort to manage, often due to young age and a lack of language skills. There is a heavy reliance on radiological findings. Whilst FVRA has previously been reported to be of value as a determinant of clinical severity in adult and paediatric patients, we find that this measurement is of dubious value in our retrospective cohort.