Introduction: The objective of this study is to evaluate the existing Spetzler-Martin grading system, and to purpose a new Cerebellar grading system to estimate the risk associated with these lesions.
Methods: Two tertiary medical centers retrospectively reviewed patients with cerebellar AVMs treated microsurgically. Patients at Institution 1 were collected from September 1999 to February 2013; at Institution 2, from October 2008 to October 2015.
Results: A total of 125 cerebellar AVMs out of 1328 Brain AVMs (BAVMs) were reviewed, 120 of which treated microsurgerically and included in the study. We found the Spetzler-Martin grading system to show no difference for outcomes between the different grades using Ficher Exact Test (p-value=0.29) or for predictability using logistic regression (p-value=0.30 AUROC=0.56). Our purposed point-based grading system includes neurologic status prior to treatment (Poor +6 Points), presurgical hemorrhage (+3 Points), age 50-60 (+1 Point) or age>60 (+3 Points), and deep venous drainage (+3 Points for Deep). We found 16.3% (8/49) of patients in Grade 1 (Points 0-4), 37.8% (14/37) in Grade 2 (Points 5-9), and 64.7% (22/34) in Grade 3 (10-15) had poor outcomes. Poor outcomes differed significantly between grades, assessed with Chi2 (p-value=<.001). Using logistic regression, Grade 2 had 3.1x the risk of experiencing a poor outcome (p-value=.03), and 9.4x the risk for Grade 3 (p-value=<.001), with an acceptable level of predictability for the model (AUROC curve=0.73).
Conclusions: We propose a novel grading system for cerebellar AVMs based on presurgical hemorrhage, venous drainage, age, and neurologic status prior to treatment. Retrospective and prospective studies with large patient samples are needed to fully establish the accuracy and predictive value of this new, proposed, grading system. Retrospective and prospective studies with large patient samples are needed to fully establish the accuracy and predictive value of this new, proposed, grading system.
Patient Care: This research will help improve patient care by providing a prognostic tool that both patients and physicians can use to allow for more informed decision making and strategic planning.
Learning Objectives: By the conclusion of this session, participants should be able to 1) describe the prognostic accuracy of the Spetzler-Martin grading system for Cerebellar AVMs, 2) identify factors having the strongest association with patient outcomes, and 3) be able to apply a novel prognostic grading system for surgically resected cerebellar AVMs.