Retroflexed dens in paediatric Chiari 1 patients and implications: Single centre retrospective study

Chiari 1 malformation and hind brain hernia can be associated with skull base and craniocervical anomalies. One of the more recently associated anomalies is a retroverted or retroflexed odontoid process or dens. We conducted a retrospective study of our operated symptomatic and conservatively managed asymptomatic Chiari cohort to assess the impact of dens retroflexion on rate of revision or cerebrospinal fluid diversion following primary foramen magnum decompression (FMD).

We undertook a retrospective study of all foramen magnum decompression (FMD) cases for Chiari type 1 malformation performed over a 15-year period in a single tertiary paediatric neurosurgical unit. For comparison, non-operated asymptomatic Chiari cases were considered as reference cohort. Information gathered included: demographics, age, sex, length of cerebellar tonsils below McRae's line, pB-C2 distance (a line drawn perpendicular to one drawn between the basion and the posterior aspect of the C2 body), angle of retroflexion (angle formed between a line drawn through the odontoid synchondrosis and its intersection with a line drawn from the tip of the odontoid process) and angle of retroversion (angle formed between the line drawn from the base of C2 and its intersection with a line drawn from the tip of the odontoid process). Grade of retroflexion was measured using pre-operative mid-sagittal MR images and classified as grade 0 (> 90°), grade 1 (85°-89°); grade 2 (80°-84°) and grade 3 (< 80°). The rates for redo surgery or need for cerebrospinal fluid (CSF) diversion were obtained from clinical records and compared in the operated and non-operated groups.

One hundred twenty-six Chiari 1 patients were included in this study with adequate imaging. Sixty-five patients were in the non-operated asymptomatic cohort with 61 patients in the operated symptomatic cohort. Mean age of non-operated cohort was 10.2 years with M:F ratio (30:35). Mean cerebellar tonsillar length below McRae's line was 10.3 mm. 7.7% of this cohort had associated syrinx. Mean angles of retroversion and retroflexion were 76 and 78°, respectively. Retroflexion grades included (9.2% grade 1, 35% grade 2 and 52.3% grade 3). pB-C2 distance was 6.8 mm. Mean age of operated cohort was 11.3 years, with M:F ratio (21:40). Mean cerebellar tonsillar length below McRae's line was 15 mm. 45.9% of this cohort had associated syrinx. Mean angles of retroversion and retroflexion were 73 and 74.5°, respectively. Retroflexion grades included (4.9% grade 1, 16.5% grade 2 and 78.6% grade 3). pB-C2 distance was 6.9 mm. No association was identified between retroflexion grade and rate of revision or CSF diversion following primary foramen magnum decompression.

The operated Chiari 1 cohort had more retroflexed dens, longer tonsils and associated syrinx compared to the non-operated asymptomatic cohort.