PURPOSE: To propose a modified angiographic classification for peripheral arteriovenous malformations (AVMs) and to determine its value for assessing therapeutic outcomes and approaches to ethanol embolization of AVMs in the body and extremities. METHODS: AVMs of the trunk and extremities were categorized according to the angiographic morphology of the nidus: type I (arteriovenous fistulae), type II (arteriolovenous fistulae), type IIIa (arteriolovenulous fistulae with non-dilated fistula), and type IIIb (arteriolovenulous fistulae with dilated fistula). Sixty-six patients (41 women; mean age 28.3 years, range 5-53) with inoperable AVMs in the body and extremities who underwent staged ethanol embolizations were retrospectively analyzed. Therapeutic outcomes and approaches were evaluated according the above classification system. RESULTS: The 66 AVMs were composed of 30 (45%) type IIIb, 13 (20%) type II, 12 (18%) mixed types IIIa and IIIb, 9 (14%) mixed types II and IIIb, and 2 (3%) type IIIa. Ethanol embolization was most effective for type II (100%), and more effective for type IIIb (83%) than for type IIIa or mixed types (< or = 50%). Despite the use of the transarterial approach, direct puncture and transvenous approaches were more relevant for treating type II AVMs. Only the transarterial approach was used for treating type IIIa; both direct puncture and transarterial approaches were used for treating the other types. CONCLUSION: The described angiographic classification provides considerable information concerning the characteristics of AVMs in the body and extremities, the optimal therapeutic approach, and the likely therapeutic outcome.
PURPOSE: To propose a modified angiographic classification for peripheral arteriovenous malformations (AVMs) and to determine its value for assessing therapeutic outcomes and approaches to ethanol embolization of AVMs in the body and extremities. METHODS: AVMs of the trunk and extremities were categorized according to the angiographic morphology of the nidus: type I (arteriovenous fistulae), type II (arteriolovenous fistulae), type IIIa (arteriolovenulous fistulae with non-dilated fistula), and type IIIb (arteriolovenulous fistulae with dilated fistula). Sixty-six patients (41 women; mean age 28.3 years, range 5-53) with inoperable AVMs in the body and extremities who underwent staged ethanol embolizations were retrospectively analyzed. Therapeutic outcomes and approaches were evaluated according the above classification system. RESULTS: The 66 AVMs were composed of 30 (45%) type IIIb, 13 (20%) type II, 12 (18%) mixed types IIIa and IIIb, 9 (14%) mixed types II and IIIb, and 2 (3%) type IIIa. Ethanol embolization was most effective for type II (100%), and more effective for type IIIb (83%) than for type IIIa or mixed types (< or = 50%). Despite the use of the transarterial approach, direct puncture and transvenous approaches were more relevant for treating type II AVMs. Only the transarterial approach was used for treating type IIIa; both direct puncture and transarterial approaches were used for treating the other types. CONCLUSION: The described angiographic classification provides considerable information concerning the characteristics of AVMs in the body and extremities, the optimal therapeutic approach, and the likely therapeutic outcome.