Spinal Pial (Type IV) Arteriovenous Fistulae

Abstract

Demographics, hemorrhage risk, and results of surgical and endovascular treatment of spinal pial (type IV) arteriovenous fistulae (AVFs) across a large patient group have not been previously reported. To report demographics, hemorrhage rates, and treatment results for these AVFs. We performed a pooled analysis via the PubMed and Embase databases through November 2012. Individualized patient data were extracted and analyzed using Cox proportional hazards regression to obtain hazard ratios for hemorrhage risk factors and pooled for baseline demographics and treatment results. We extracted information on 213 patients from 28 studies. Only 1% of lesions were incidental; 93% of patients presented with neurologic deficits and 36% with hemorrhage. Patients with type IVa lesions were significantly older (mean age, 46.9 years) and demonstrated a male sex predilection (68% male). Patients with type IVc lesions were significantly younger (mean age, 18.7 years), had no sex predilection, and had the highest prevalence of syndromic conditions (29% of cases). The annual hemorrhage rate was 2.5% (95% confidence interval [CI]: 1.4%-4.7%), increasing to 5.6% for hemorrhagic fistulae (95% CI: 3.0%-10.7%; hazard ratio: 6.31; 95% CI: 0.69-57.4; P = .10). Patient sex, fistula location, and fistula subclass were not significant risk factors for hemorrhage. The surgical obliteration rate was 88%; 68% of patients improved, 26% were the same, and 6% were worse. The endovascular obliteration rate was 74%; 75% of patients improved, 14% were the same, and 11% were worse. We demonstrate the utility of the Anson-Spetzler a-c subclassification and underscore the efficacy of surgical and endovascular spinal AVF treatment.