BACKGROUND: The natural history and treatment results for spinal glomus (type II) arteriovenous malformations (AVMs) remain relatively obscure.
OBJECTIVE: To calculate spinal glomus (type II) AVM hemorrhages rates and amalgamate results of intervention.
METHODS: We performed a pooled analysis via the PubMed database through May 2012, including studies with at least 3 cases. Data on individual patients were extracted and analyzed using a Cox proportional hazards regression model to obtain hazard ratios for hemorrhage risk factors.
RESULTS: The annual hemorrhage rate before treatment was 4% (95% confidence interval [confidence interval]: 3%-6%), increasing to 10% (95% CI: 7%-16%) for AVMs with previous hemorrhage. The hazard ratio for hemorrhage after hemorrhagic presentation was 2.25 (95% CI: 0.71-7.07), increasing to 13.0 within the first 10 years (95% CI: 1.44-118). The overall rates of complete obliteration were 78% (95% CI: 72%-83%) for surgery and 33% (95% CI: 24%-43%) for endovascular treatment. Long-term clinical worsening occurred in 12% of patients after surgical treatment (95% CI: 8%-16%) and in 13% after endovascular treatment (95% CI: 7%-21%). No hemorrhages occurred after complete obliteration. After partial surgical treatment, the annual hemorrhage rate was 3% (95% CI: 1%-6%); no hemorrhages were reported over 196 patient-years after partial endovascular treatment.
CONCLUSION: Spinal glomus (type II) AVMs with previous hemorrhage, particularly within 10 years, demonstrated a greater risk of hemorrhage. Complete obliteration and even partial endovascular treatment significantly decreased their hemorrhage rate.
ABBREVIATIONS: AVM, arteriovenous malformation
CI, confidence interval
NBCA, N-butyl cyanoacrylate
PVA, polyvinyl alcohol