systemic circulation," thereby producing serious occlusion of a systemic artery. Detachable silicone balloons and stainless steel coils with Dacron fibers have been used for 17 years by interventional radiologists to close PAVMs, and my colleagues and I believe that both have an important role in providing safe and efficacious treatment.2 We have not observed late migration of a detachable silicone balloon or a coil once they are securely in place in the artery leading to a PAVM in any of the more than 250 patients whom we have treated. We do agree that balloons, like coils, can dislocate at the time of place¬ ment if they are not properly secured into the artery lead¬ ing to the PAVM. If migration of the occlusion device oc¬ curs at the time of placement, it will enter the systemic circulation with potentially serious consequences.2The experience with detachable silicone balloons, both experimentally and clinically, demonstrates that when they deflate, and approximately 10% do over the course of the first 1 to 2 years, the deflated shell will remain attached to the wall of the artery or embedded within an organized thrombus.3*4 Experimentally, balloons deflating even af¬ ter as short a time as 2 weeks after placement are also trapped within an organized thrombus.3 The late defla¬ tion that Haitjema and colleagues reference in the series of patients described by Puskas et aL involved 1 patient with a deflation of f balloon documented at 4 months af¬ ter placement. There was no clinical evidence mentioned to indicate that the balloon or its deflated shell migrated into the systemic circulation. The angiograms from this patient were not shown. It is quite possible that, rather than recanalization, this PAVM recurred because an accessory vessel had enlarged at the time of restudy 5 years later, ac¬ counting for the continued patency.Detachable silicone balloons filled with isosmotic contrast media produce excellent cross-sectional occlu¬ sion of PAVMs.4*6*7 My colleagues and I believe that there are advantages and disadvantages to the use of both coils and balloons, depending on the anatomical situation, and that radiologists treating PAVMs should have access and familiarity with both types of occlusion devices. Robert I. White, Jr, MD New Haven, Conn 1. Haitjema T, Westermann CJJ, Overtoom TTC, et al. Hereditary hemorrhagic telangiectasia (Osler-Weber-Rendu disease): new insights in pathogenesis, complications, and treatment. Arch Intern Med. 1996;156:714-719. 2. White RI, Nyhan AL, Terry P, Buescher PC, et al. Pulmonary arteriovenous malformations: techniques and long-term outcome of embolotherapy. Radiology. 1988;169:633-669. 3. Kaufman SL, Strandberg JV, Barth KH, Gross GS, White RI. Therapeutic embolization with detachable Silastic balloon: long-term effects in swine. Invest Radiol. 1979;14:156-161. 4. Pollak JS, Egglin TK, Rosenblatt MM, Dickey KW, White RI. Clinical results of transvenous systemic embolotherapy with a neuroradiological detachable balloon. Radiology. 1994;191:477-482. 5. Puskas JD, Allen MS, M...