Skip to main content
  • Outcome Predictors of Brainstem Cavernous Malformations

    Final Number:

    Roxanna Garcia; Michael Ivan; Michael T. Lawton MD

    Study Design:

    Subject Category:

    Meeting: Congress of Neurological Surgeons 2013 Annual Meeting

    Introduction: In this retrospective review the authors evaluate the outcomes of patients with brainstem cavernous malformations (BSCMs) treated surgically. We hypothesized that pre-operative outcome predictors for BSCMs are similar to those used in arteriovenous malformations, including the Spetzler-Martin criteria and supplementary grading.

    Methods: A consecutive series of 104 patients from 1997-2012 was used to assess preoperative clinical predictors associated with post-surgical patient outcomes. Univariable logistic regression was performed to identify predictors and a multivariable logistic regression model was constructed to test the association of the combined predictors with the final Modified Rankin Scale (mRS) scores. The multivariable models were 10-fold cross-validated to determine the misclassification rate. A grading system assigned points for age, lesion size, presence of developmental venous anomaly (DVA), brainstem location, and time since last hemorrhagic event.

    Results: The mean age was 42.1 years (range 7-81 years) and a mean follow-up period of 18.6 months. The average lesion size was 19.3mm. Approximately, 89.5% of patients either showed improved outcomes (53%) by change in pre-operative to final post-operative mRS score or did not worsen from their preoperative status (36.5%). The overall morbidity including motor, cranial nerve and ataxia was 14.4% and there was one mortality. Significant preoperative predictors identified by univariate analysis. A multivariable model adjusted for follow-up time included all significant variables and showed ROC area of 0.87 (95% CI: 0.80 - 0.94). The final model included age, lesion width, DVA presence, crossing of the axial midpoint of the brainstem, and time since last hemorrhagic event. A 10-fold cross-validation of the full model estimated a 17.1% (95% CI: 17.0-17.2) misclassification rate.

    Conclusions: Brainstem cavernous malformations likely have similar predictive factors compared to arteriovenous malformations. Although the small sample size limits our statistical power, we were able to identify new predictive factors that can aid in assessing patient outcomes following surgical resection of brainstem cavernous malformations.

    Patient Care: Allow surgeons to use preoperative clinical factors to assist in predicting outcomes of surgically treated brainstem cavernous malformations

    Learning Objectives: To use preoperative clinical factors to assist in predicting outcomes of surgically treated brainstem cavernous malformations


We use cookies to improve the performance of our site, to analyze the traffic to our site, and to personalize your experience of the site. You can control cookies through your browser settings. Please find more information on the cookies used on our site. Privacy Policy