Validation of fourth ventricle roof angle as a measure of brainstem dysfunction in pediatric Chiari malformation type...

Chiari malformation type I (CM-I) is a common condition characterized by the cerebellar tonsillar position below the foramen magnum. Radiographic measurements are important for diagnosing CM-I but have limited use in predicting the severity of the condition and the need for surgery. The fourth ventricle roof angle (FVRA), a measure of deformation or bowing of the roof of the fourth ventricle, has been presented as a tool for predicting brainstem dysfunction in patients with CM-I; however, this has not been validated. Utilizing a database of pediatric patients with CM-I from a large single center, this study aimed to validate the finding that the FVRA can serve as a predictor of brainstem dysfunction in pediatric CM-I and to present additional potential correlations of interest between the FVRA and symptoms and other radiographical measurements.

Radiographic measurements were retrospectively reviewed for 388 pediatric patients evaluated for CM-I at Children's of Alabama from November 2010 to 2017. Fourth ventricle bowing was determined to be present if the FVRA was > 65°. Univariate and multivariate logistic regression analyses were performed to identify significant associations with the FVRA, brainstem dysfunction, and surgery. Brainstem dysfunction was defined as the presence of lower cranial nerve dysfunction and/or sleep apnea.

Multivariate logistic regression showed significant association of lower brainstem dysfunction with tonsillar position (OR 1.17, 95% CI 1.06-1.29), basilar invagination (OR 0.66, 95% CI 0.47-0.91), supraoccipital length (OR 0.86, 95% CI 0.77-0.96), and tussive headache (OR 6.62, 95% CI 1.95-22.46). The presence of bowing did not show a significant association with brainstem dysfunction in the multivariable model (OR 1.02, 95% CI 0.27-3.81), and it was not significantly associated with undergoing surgery after controlling for symptoms that always resulted in surgery (p = 0.483).

Univariate logistic regression analysis demonstrated that an FVRA greater than 65° is significantly associated with brainstem dysfunction. In a multivariate logistic regression model, however, the presence of bowing was not independently associated with brainstem dysfunction. As such, additional studies are needed before FVRA can be utilized clinically.