Nongalenic Arteriovenous Fistulas: History of Treatment and Technology

Kristen Upchurch, MD; Lei Feng, MD, PhD; Gary R. Duckwiler, MD; John G. Frazee, MD; Neil A. Martin, MD; Fernando Viñuela, MD


Neurosurg Focus. 2006;20(6) 

In This Article

History of Surgical Treatment: Dandy and Drake

The goal of nongalenic AVF treatment is obliteration of the fistula via occlusion of its arterial supply without premature occlusion of its venous drainage.[5,17,35,37] This objective can be achieved using surgical or endovascular therapy, separately or in combination.[3,21,36,38] Historically, nongalenic AVFs were treated at craniotomy by occluding feeding arteries as close as possible to the site of the fistula, initially by ligation and later by clipping.[5,8,11] A review of cases of nongalenic cerebral AVF reported between 1928 and the present re veals that the original role of surgery was augmented over time by complementary or alternative endovascular treatment, due to technological progress in interventional neuroradiology.

Surgical Treatment by Dandy

Open surgery with ligation was at first the only treatment option for nongalenic AVFs. In his 1928 series of "arterio venous aneurysms of the brain," Dandy[8] reported the surgical treatment of a lesion that he described as a "pure arterio venous fistula," that is, a nongalenic AVF supplied by a single feeding artery off the right MCA, which connected directly to an enlarged venous trunk (Fig. 7). At the initial operation, Dandy performed a craniotomy, observed the massive venous drainage of the lesion on the sur face of the right temporal and frontal lobes, and then closed. He wrote, "Since any attempt to treat the aneurysm surgically appeared to involve a risk greater than my co-workers and I were justified in assuming, the dura was closed and the bone flap was replaced." At reoperation, the feeding artery was ligated and the varix was then observed to collapse immediately (Fig. 8). Although the 52-year-old patient's seizure disorder was cured by the surgical treatment of his AVF, postoperatively he had a new permanent deficit of mild left hemiparesis. Dandy reached a cautionary conclusion in summarizing his clinical experience with this lesion and other types of cerebral arteriovenous aneurysms. He wrote, "Radical ligations or extirpations alone are curative, but are exceedingly dangerous to life and function and are indicated in the minority of cases. . . ." Dandy's surgical treatment of these cerebrovascular lesions was limited by the technology available in his time (Fig. 9 left).

Figure 7.

Sketch by Dandy of a nongalenic AVF that he treated surgically, as described in his 1928 series. The lesion was supplied by a single feeding artery off the right MCA that emptied directly into a varix. a. = artery; br. = branch; mid. cerebr. = middle cerebral.

Figure 8.

Dandy's illustration of the same nongalenic AVF shown in Fig. 7, observed in situ at craniotomy, demonstrating ligation of the feeding artery.

Figure 9.

Photographs of W. E. Dandy, M.D. (left) and C. G. Drake, M.D. (right).

Surgical Treatment by Drake

In the 1960s and 1970s, technological innovations led to intraoperative and transfemoral embolization as a new strategy for treating cerebrovascular lesions. Yet open surgery with clip occlusion of feeding arteries remained the treatment option of choice for most cerebral AVMs and invariably also for nongalenic AVFs in the 1970s. In his 1979 series of 166 patients with cerebral AVMs, Charles Drake (Fig. 9 right) included two cases of nongalenic AVFs associated with huge varices. Although five AVMs in this series were treated with flow-directed or direct selective plastic bead embolization, this type of procedure was not an option for nongalenic AVFs, given the dangers posed by the AVFs' high flow and associated massive varix. The two patients with nongalenic AVFs in Drake's series underwent surgical clip occlusion of feeding arteries. This was a success in one patient, but the other patient died 24 hours after the operation due to intracerebral hemorrhage attributed to hemodynamic alteration. In this case of death resulting from hemorrhage, the venous outlet of the AVF was not surgically occluded during the operation, and later autopsy confirmed the patency of the varix. As in the case of Dandy's patient, the varix collapsed on itself after Drake occluded the solitary large feeding artery of the AVF (in this case with a single clip). Drake[11] hypothesized that the cause of hemorrhage was rupture of a deep normal artery.

Surgical treatment of nongalenic AVFs poses some disadvantages particular to the lesions' anatomy. For example, at craniotomy, visualization and access to the fistula itself are often hindered by the presence of a massive associated varix.[8,17,25] Endovascular therapy avoids this problem and is now a frequently selected treatment option for nongalenic AVFs. The modern history of the innovations underlying the optimization of endovascular treatment for nongalenic AVFs is discussed in the following section.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: