Intraoperative neuromonitoring for pediatric Chiari decompression: when is it useful?

Surgical treatment for symptomatic Chiari I malformation involves surgical decompression of the craniovertebral junction. Given the proximity of critical brainstem structures, intraoperative neuromonitoring (IONM) is employed for safe decompression in some institutions. However, IONM adds time and cost to the operation, and the benefit to the patient has not been defined. Given the diversity in surgical practices, there is no evidence-based standard of care regarding when to use IONM and which modalities are most helpful. The purpose of this study was to review a single-surgeon experience with IONM in order to determine the sensitivity, specificity, and predictive values of various IONM modalities routinely used in pediatric Chiari I decompression; to examine the associations between patient, clinical, and radiographic characteristics and IONM alerts; and to obtain data regarding the usefulness of these modalities during the surgical process to improve patient outcomes.

A retrospective review was performed for 300 consecutive pediatric patients who underwent suboccipital craniectomy and C1 laminectomy for Chiari decompression performed by a single surgeon over a 15-year period. Clinical, radiographic, and IONM data were collected. Radiographic measurements of the skull base morphological abnormalities, including clival angle, Chamberlain's line, and Grabb-Oakes line, were compared between patients with and without true IONM signal changes.

A total of 291 cases were included, with an age range of 6 months to 19 years. Among 291 cases, somatosensory evoked potentials (SSEPs) were monitored in 291, motor evoked potentials (MEPs) in 209, cranial nerve spontaneous electromyography (sEMG) in 290, and brainstem auditory evoked potentials (BAEPs) in 110. Sensitivity, specificity, positive predictive value, and negative predictive value, respectively, were as follows: 1.00, 1.00, 1.00, and 1.00 for SSEPs; 1.00, 0.99, 0.67, and 1.00 for MEPs; 0.00, 0.88, 0.00, and 1.00 for sEMG; and not appliable, 1.00, not applicable, and 1.00 for BAEPs. Six patients had true IONM signal changes. These patients had radiographic evidence of more severe concomitant craniocervical instability and basilar invagination, with steeper clival angles (124° vs 146°, p = 0.02) and larger Grabb-Oakes lines (10.1 mm vs 6.7 mm, p = 0.02), when compared with the patients without any true IONM changes.

Intraoperative neuromonitoring may be best utilized for patients who show radiographic features of abnormal skull base morphology, defined as a clival angle < 135° or Grabb-Oakes line > 9 mm. When IONM is employed, SSEP and MEP monitoring are the most useful modalities.