Garcia Roxanna M, Ivan Michael E, Lawton Michael T
*Department of Neurological Surgery, and ‡Center for Cerebrovascular Research, University of California at San Francisco, San Francisco, California.
Neurosurgery. 2015 Mar;76(3):265-77; discussion 277-8. doi: 10.1227/NEU.0000000000000602.
Once considered inoperable lesions in inviolable territory, brainstem cavernous malformations (BSCM) are now surgically curable with acceptable operative morbidity. Recommending surgery is a difficult decision that would be facilitated by a grading system designed specifically for BSCMs that predicted surgical outcomes.
Informed by our efforts to develop a supplementary grading system for arteriovenous malformations, we hypothesized that a similar system might predict long-term outcomes and guide clinical decision-making.
A consecutive, single-surgeon series of 104 patients was used to assess preoperative clinical and imaging predictors of microsurgical outcomes. Univariable logistic regression identified predictors and a multivariable logistic regression model tested the association of the combined predictors with final modified Rankin Scale scores. A grading system assigned points for lesion size, location crossing the brainstem's midpoint, presence of developmental venous anomaly, age, and time from last hemorrhage to surgery.
Average maximal diameter of BSCMs was 19.5 mm; 50% crossed the axial midpoint; 54.8% had developmental venous anomalies; mean age was 42.1 years; and median time from last hemorrhage to surgery was 60 days. One patient died (0.96%), and 15 patients (14.4%) experienced worsened cranial nerve or motor dysfunction, of which 10 increased their modified Rankin Scale scores (9.6%). BSCM grades ranged from 0 to 7 points and predicted outcomes with high accuracy (receiver operating characteristic = 0.86, 95% confidence interval: 0.78-0.94).
Rather than developing a grading system for all cerebral cavernous malformations that is weak with BSCMs, we propose a system for the patients who need it most. The BSCM grading system differentiates patients who might expect favorable surgical outcomes and offers guidance to neurosurgeons forced to select these patients.
脑干海绵状血管畸形（BSCM）曾被认为是位于不可侵犯区域的无法手术切除的病变，如今手术可治愈且手术发病率可接受。推荐手术是一个艰难的决定，而专门为BSCM设计的、能预测手术结果的分级系统将有助于做出此决定。
基于我们为动静脉畸形开发补充分级系统的工作，我们假设类似的系统可能预测长期结果并指导临床决策。
采用连续的、由单一外科医生治疗的104例患者系列，评估显微手术结果的术前临床和影像预测因素。单变量逻辑回归确定预测因素，多变量逻辑回归模型测试组合预测因素与最终改良Rankin量表评分之间的关联。分级系统根据病变大小、跨越脑干中点的位置、发育性静脉异常的存在、年龄以及从上一次出血到手术的时间来评分。
BSCM的平均最大直径为19.5毫米；50%跨越轴位中点；54.8%有发育性静脉异常；平均年龄为42.1岁；从上一次出血到手术的中位时间为60天。1例患者死亡（0.96%），15例患者（14.4%）出现颅神经或运动功能障碍恶化，其中10例患者改良Rankin量表评分增加（9.6%）。BSCM分级范围为0至7分，预测结果的准确性较高（受试者工作特征曲线下面积=0.86，95%置信区间：0.78-0.94）。
我们不是为所有脑海绵状血管畸形开发一个对BSCM效果不佳的分级系统，而是为最需要的患者提出一个系统。BSCM分级系统区分了可能预期手术结果良好的患者，并为被迫选择这些患者的神经外科医生提供指导。
