External validation of the Lawton brainstem cavernous malformation grading system in a cohort of 277 microsurgical patients.

ABSTRACT

OBJECTIVE: The brainstem cavernous malformation (BSCM) grading system predicts neurological outcomes associated with microsurgical resection and assists neurosurgeons in selecting patients for treatment. The predictive accuracy of the BSCM grading system should be validated in a large cohort from high-volume centers to generalize its use. METHODS: An external validation cohort comprised patients with a BSCM resected by the senior author (M.T.L.) since the publication of the BSCM grading system and those resected by another neurosurgeon (R.F.S.) over a 16-year period. Size, crossing the axial midpoint, the presence of a developmental venous anomaly, patient age, and timing of last hemorrhage were used to assign BSCM grades from 0 to VII. Poor neurological outcomes were recorded as modified Rankin Scale scores > 2 at last follow-up examination. RESULTS: A total of 277 patients were included in the study. The average BSCM grade was 3.9, and the majority of BSCMs (181 patients, 65%) were intermediate grade (grades III-V). Outcomes were predicted by BSCM grade, with good outcomes observed in 47 of 54 patients (87%) with low-grade BSCMs, in 135 of 181 patients (75%) with intermediate-grade BSCMs, and in 21 of 42 patients (50%) with high-grade BSCMs. Conversely, proportions of patients with neurological deterioration increased with increasing BSCM grade, with worsening observed in 2 of 54 patients (4%) with low-grade BSCMs, in 29 of 181 patients (16%) with intermediate-grade BSCMs, and in 17 of 42 patients (40%) with high-grade BSCMs. In the chi-square analysis, high-grade BSCMs were associated with increased odds of neurological worsening compared to low- and intermediate-grade BSCMs (OR 5.0, 95% CI 2.4-10.4; p < 0.001). The receiver operating characteristic analysis demonstrated acceptable discrimination for predicting unfavorable functional outcomes (modified Rankin Scale score > 2) with an area under the curve of 0.74 (95% CI 0.68-0.80; p < 0.001). CONCLUSIONS: This study validates the BSCM grading system in a large cohort of patients from two high-volume surgeons. BSCM grade predicted neurological outcomes with accuracy comparable to that of other grading systems in widespread use. The BSCM grading system establishes categories of low-, intermediate-, and high-grade BSCMs and a boundary or cutoff for surgery at BSCM grade V. BSCM grading guides the analysis of a particular patient's condition, but treatment recommendations must be individualized, and neurosurgeons must calibrate BSCM grading to their own outcome results, unique abilities, and practices.