Aneurysms of the ASA and its branches are extremely rare and they may occur as isolated lesions without an identifiable reason [
1‐
3]. The assumption of an underlying dissection sometimes appears to be rather speculative [
3,
4]. Cervical spondylosis, per se a very frequent disorder, has been reported to be associated with the formation of an adjacent ASA aneurysm [
5].
As in the case reported here, ASA aneurysms can be found in conjunction with a variety of arteriovenous shunts. They are usually not encountered in spinal arteriovenous fistulas. Biondi et al. (1992) [
6,
7] as well as Konan et al. [
8] described sulco-commissural aneurysms, which were located on feeding arteries of spinal cord arteriovenous malformations. In these cases both the hemodynamic stress due to the increased blood flow as well as a developmental factor may have contributed to the aneurysm formation. In ASA aneurysms associated with steno-occlusive disease of major cranial arteries, the collateral flow via the spinal vasculature is the most likely cause of aneurysm development [
9‐
12]. More complex etiopathogenetic relationships become apparent in ASA aneurysms in the context of epidural and intradural arteriovenous fistulas. Takasaki et al. (1991) [
13] clipped an ASA aneurysm in a patient with a dural arteriovenous fistula of the posterior fossa. In this case, due to the poor image quality of the available prints, it is difficult to determine if the ASA contributed to the dural arteriovenous shunt. Most likely there was no hemodynamic connection between the fistula and the aneurysm (Ryogo Anei, Department of Neurosurgery, Asahikawa Medical University, Hokkaido, Japan; personal communication). In the patient reported by Gilard et al. (2013) [
14] the dural arteriovenous fistula was located at the foramen magnum and the ASA with the aneurysm supplied the fistula through the left C1 radiculopial artery. Both patients presented with an SAH. Matsui et al. (2007) [
15] and Onda et al. (2012) [
16] operated on patients presenting with hematomyelia due to combined dural AV shunts and spinal artery aneurysms. The AV shunt was located at the C1/C2 level and was supplied by both dural and pial feeding arteries, including the ASA. The aneurysms in both cases most likely arose from a sulco-commissural artery, because they were located lateral to the expected continuation of the ASA. Since these arteries participated in the supply of the AV shunt the aneurysms can be considered as flow-related. A patient presenting with an SAH and otherwise features almost identical to the present case was reported by Nakagawa et al. (2014) [
17]. An AV fistula located at C1–C2, which was called “epidural” by the authors, was supplied from the radicular artery and from pial vessels arising from the ASA. The authors used the term “anterior spinal artery aneurysm”. The DSA images, however, show the aneurysm lateral to the expected continuation of the ASA, as would be the case in an aneurysm of the sulco-commissural artery. In the present patient, an intradural AV fistula on the right side at the C2 level was supplied by the radicular artery from the right vertebral artery as well as from several pial arteries originating from the ASA. The venous drainage was exclusively paraspinal. The assumption was that the intradural AV fistula with the radicular supply was the primary lesion. It appears at least possible that such a fistula, located
in the dura, may recruit supply from pial vessels. A flow-related induction of the aneurysm formation seems plausible. Alternatively, a metameric or myelomeric connection between the AV shunt and the aneurysm is at least a possibility [
18]. The documented growth of the aneurysms is an interesting parallel between the case of Nakagawa et al. (2014) [
17] and the present case.
The management of this unusual condition is a matter of controversy and the clinical decision making must take the patient’s individual circumstances into account. A poor clinical condition or advanced age are certainly arguments in favor of conservative management, and a good outcome is possible [
2,
5], even with spontaneous thrombosis of the aneurysm [
3,
19]. On the other hand, re-hemorrhage and death may occur [
20]. Direct surgical treatment of such aneurysms has been infrequently reported [
9,
13,
15,
21,
22]. The
indirect influence on ASA aneurysms by interrupting the AV shunt [
16] or placing a flow diverter in the vertebral artery covering the origin of the ASA [
23] is an interesting concept.
The aneurysm in the present patient was not located in continuity with the ASA but originated from a sulco-commissural artery, the small caliber of which made aneurysm catheterization impossible. Any further observation was discarded after the aneurysm had grown considerably over 5 months of conservative management. A transarterial embolization of the intradural arteriovenous fistula would have been technically straightforward. Obliterating an AV shunt in order to induce the shrinkage of a flow-related aneurysm is a popular concept in the treatment of pial brain AVMs. Complete proximal aneurysm obliteration is, however, more the exception than the rule [
24]. The hemodynamic strategy was dismissed since an interruption of the dural supply without impact on the pial feeding arteries might have even increased the hemodynamic stress on the ASA and the unsecured aneurysm. The glial scar around the aneurysm, resulting from the previous hemorrhage, enabled the dissection of the cervical spinal cord without further damage. It was also possible to perform a resection of this aneurysm as it was separate from the ASA. The previous hemorrhage facilitated the microsurgical extirpation without further damage to the spinal cord. A microsurgical obliteration of the epidural AV fistula was attempted but was only partially successful. An embolization of the remaining dural supply of the epidural AV fistula was offered to the patient, who was hesitant to undergo further treatment for the time being.
The poor clinical outcome (mRS 4) of this patient might be an argument in favor of pro-active treatment in patients diagnosed prior to hemorrhage.