Effect of presenting hemorrhage on outcome after microsurgical resection of brain arteriovenous malformations

OBJECTIVE: We hypothesized that patients with unruptured arteriovenous malformations (AVMs) at presentation have an increased risk of deterioration compared with patients with ruptured AVMs. METHODS: A consecutive series of 224 patients treated microsurgically by a single neurosurgeon during a period of 6.4 years was analyzed. Initial hemorrhagic presentation was the primary predictor variable. Neurological outcomes were assessed by use of the Modified Rankin Scale (MRS) and Glasgow Outcome Scale (GOS), and logistic regression identified predictors of deterioration at follow-up (mean duration, 1.3 yr) relative to baseline before any intervention. RESULTS: Overall, 120 patients (54%) presented with hemorrhage, and all 224 patients underwent microsurgical resection. Complete resection was achieved in 220 patients (98%). According to GOS score, 13 patients (5.8%) deteriorated; according to MRS score, 45 patients (20.1%) deteriorated. Fifteen patients (6.7%) died. Hemorrhagic presentation was associated with improved outcomes, with a mean change in MRS score of +0.89 in patients with ruptured AVMs and -0.38 in patients with unruptured AVMs (P \u3c 0.001). The final mean MRS scores in patients with unruptured AVMs were better than those in patients with ruptured AVMs (1.44 versus 1.90; P = 0.048). Presentation with an unruptured AVM was a predictor of worsening MRS score (odds ratio, 2.33; 95% confidence interval, 1.3-4.3; P = 0.006) but not of worsening GOS score. CONCLUSION: Presentation with AVM hemorrhage is an underappreciated predictor of outcome after therapy that includes microsurgical resection. Patients with ruptured AVMs tended to have deficits at presentation and generally improved after surgery, whereas patients with unruptured AVMs tended to have normal or nearly normal neurological function at presentation and were susceptible to worsening, albeit slight, as measured by MRS scores. Sensitive outcome measures such as MRS detect subtle symptoms and impairments missed by coarser measures such as GOS. Patients should be counseled that the risks associated with elective resection of unruptured AVMs may be higher than recognized previously. Hemorrhagic brain injury and its secondary effects may mask this surgical morbidity