- Email: [email protected]
Spinal dural arteriovenous fistula and venous aneurysm: an unusual association
Vascular
Spinal Dural Arteriovenous Fistula and Venous Aneurysm: An Unusual Association ¨ zgen, M.D. Servet Inci, M.D., Ph.D, Pınar Akdemir, M.D., and Tuncalp O Department of Neurosurgery, School of Medicine, University of Hacettepe, Ankara, Turkey
¨ zgen T. Spinal dural arteriovenous fistula and Inci S, Akdemir P, O venous aneurysm: an unusual association. Surg Neurol 2003;60: 334 – 8. OBJECTIVE
Although spinal intradural arteriovenous malformations have frequently been associated with venous aneurysms, spinal dural arteriovenous fistulas typically are not. We describe a case of conus medullaris compression by a large saccular venous aneurysm of a spinal dural arteriovenous fistula. As these lesions are slow-flow arteriovenous shunts, development of a venous aneurysm is an unexpected condition. Pathogenesis of these aneurysms is briefly discussed. METHODS
A 61-year-old man presented with progressive spastic paraparesis of 1-year duration. Spinal magnetic resonance (MR) suggested abnormal serpiginous vessels and a partially thrombosed aneurysm at the level of conus medullaris. Spinal angiography was performed but some lower thoracal and upper lumbar segments could not be catheterized because of severe atherosclerosis. As a result, vascular anomaly could not be demonstrated angiographically, but surgical exploration was found to be necessary because of positive clinical and MR findings. RESULTS
An arteriovenous fistula that entered the spinal canal beneath the L1 pedicle on the left was identified. It was easily coagulated and cut. The partially thrombosed aneurysm within the conus medullaris was also resected. CONCLUSION
Our case illustrates two important points related to patients harboring spinal dural arteriovenous fistulas (AVF): First, clinical suspicion and MR findings are important in diagnosis of these lesions and surgical exploration is indicated despite negative or nondiagnostic angiographic results in such cases. Second, venous aneurysms may be associated with spinal dural AVFs, and although spinal dural AVFs are low-flow lesions, the development of the venous aneurysm is probably a result of high venous pressure. © 2003 Elsevier Inc. All rights reserved. KEY WORDS
Spinal dural AVF, venous aneurysm, conus medullaris.
Address reprint requests to: Dr. Servet Inci, Emek Mahallesi, 4. Cadde, No:70/8, 06510 Ankara, Turkey. Received July 19, 2002; accepted December 30, 2003. 0090-3019/03/$–see front matter doi:10.1016/S0090-3019(03)00239-8
pinal dural arteriovenous fistulas (AVF) are abnormal communications between radicular arteries and perimedullary veins or coronal venous plexus [4,13]. The fistula is usually located within the dura covering of the spinal nerve root. These lesions, which are the most common type of spinal vascular malformations in adults, cause a gradually progressive myelopathy as a result of venous hypertension in the spinal cord. Although intradural spinal arteriovenous malformations are frequently associated with venous aneurysms [8,15], spinal dural AVFs are typically not associated with venous aneurysms in the large series [16-20]. We report a case of typical saccular venous aneurysm presenting with conus medullaris compression.
S
Case Reports A 61-year-old man presented with progressive spastic paraparesis of 1-year duration. The patient was evaluated in a local hospital and was referred to our department for diagnosis and treatment. Neurologic examination revealed severe weakness of both legs involving the proximal and distal muscles. He was not ambulatory and had urinary retention, decreased anal spinchter tone, and decreased sensation below T10 level. Spinal magnetic resonance (MR) showed multiple punctuate and serpiginous areas of low intensity on the dorsal surface of the spinal cord and conus medullaris. In addition, there was a rounded hypoisointense lesion at the level of conus on T2weighted sections, suggesting a thrombosed aneurysm (Figure 1). Another typical finding on the same image was increased signal intensity within the cord from T11 to the end of conus medullaris. A spinal vascular anomaly was considered. Spinal angiography was performed through the femoral route, but some lower thoracal and upper © 2003 Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010 –1710
Spinal AVF and Venous Aneurysm
Surg Neurol 335 2003;60:334 – 8
2
Histologic appearance of the partially thrombosed aneurysm.
lumbar segments, including the artery of Adamkiewicz, could not be catheterized because of severe atherosclerosis. So vascular anomaly could not be demonstrated angiographically, but surgical exploration was found to be necessary because of positive clinical and MR findings.
was extended to L1. An arteriovenous fistula that entered the spinal canal beneath the L1 pedicle on the left was identified. It coursed from the inner surface of the dura to communicate with the dilated, tortuous vein. The fistula was dissected away from nerve root and easily coagulated and cut. The draining vein gradually collapsed within several minutes. Just as the vein collapsed, a rounded, dark blue, cystic lesion was noticed on the conus medullaris. This lesion was just distal to the fistula and buried in the conus medullaris. The well-circumscribed, nut-sized lesion was separated from the spinal cord with gentle dissection. The dome of the lesion was incised, with escape of chocolate-colored fluid. It was a partially thrombosed aneurysm that was resected because of mass effect. It was histologically verified as a partially thrombosed aneurysm (Figure 2).
OPERATION The thoracolumbar region from T9 to L2, because some of these segments could not be visualized on preoperative angiography, was cleaned and draped. Initially, T11 and T12 laminectomies were carried out. The dura was opened longitudinally under the microscope, and a dilated, tortuous vein was encountered on the dorsal surface of the cord. A systematic intradural exploration of each radicular artery at nerve root level was bilaterally examined, but a fistula could not be found. The artery of Adamkiewicz was also identified at left T12 level as normal caliber and appearance. Then laminectomy
POSTOPERATIVE COURSE Despite some initial improvement, the patient had no long-term change in his neurologic deficits. His long-standing lower extremity spasticity did not resolve but has not progressed over the subsequent 6 months. Control spinal angiography was not performed because AVF could not be demonstrated preoperatively. Control spinal MR was performed on postoperative 10th day, and T2-weighted sections showed no tortuous vessels or aneurysm. Furthermore, increased signal intensity within the spinal cord had been partially decreased (Figure 3).
Sagittal T2-weighted spinal MR demonstrates serpiginous signal void areas on the dorsal surface of the cord, increased signal intensity within the spinal cord from T11 to L2 and an intramedullary rounded mass (arrow) in the conus medullaris. At surgery, the mass was found to be a partially thrombosed venous aneurysm distal to the fistula.
1
336 Surg Neurol 2003;60:334 – 8
Postoperative sagittal T2-weighted spinal MR shows no tortuous vessels. Increased signal intensity within the spinal cord is mostly resolved.
3
Discussion Spinal DAVF was first described by Kendall and Louge [10] as “spinal epidural angiomatous malformations draining into intrathecal veins,” in 1977. Three years later, Merland et al [14] reported these lesions as “radiculomeningeal arteriovenous fistula.” Spinal dural AVF (also known as Type I arteriovenous malformation) is the most common spinal vascular malformation, representing more than two-thirds of all types [5,18]. They are usually found in the thoracolumbar region in middle-aged men [3,16,18], as in our case. Their pathogenesis is still debatable. However, they are considered to be acquired, probably as a result of venous thrombosis or trauma [13]. Such a fistula consists of a single arterial feeder (occasionally two or three feeders) from a radicular artery that enters the dura at the dural root sleeve. Spinal cord arteries never participate in the shunt. This is the most important difference of dural AVF from spinal cord AVF. The shunt is drained intradurally by retrograde flow through medullary vein, resulting in engorgement of the coronal venous plexus and intraparenchymal radial veins [17,18]. It is now well accepted that this venous hypertension accounts for the progressive
Inci et al
myelopathy. The arteriovenous fistula causes a chronic venous congestion of the spinal cord and secondary ischemia [2,9], which is responsible for the clinical picture of myelopathy. Symptoms are usually gradually progressive, consisting of weakness and sensory symptoms in the legs with a severe disturbance in micturition. Spinal angiography, of course, is the gold standard for the diagnosis of spinal dural AVFs. Our angiography was incomplete because of technical difficulties due to severe atherosclerosis. Rarely, arteriovenous fistulas cannot be demonstrated because of some other reasons: (1) The fistula itself can be microscopic [12]; (2) the shunt flow can be extremely slow [19]. Recently, Alleyne et al [1] also reported three patients with angiographically occult spinal dural AVFs. In these situations, a high index of clinical suspicion should be key in diagnosing these lesions, and surgical exploration is indicated despite negative angiographic results. Sometimes the location of the fistula may not correlate with the level and site of the clinical symptoms [17,18]. Therefore, the patient should be prepared for a large exploration. Spinal dural AVFs are frequently diagnosed after delays because they are usually not included in the differential diagnosis of myelopathy. Sometimes, they are also misdiagnosed as transverse myelitis, intramedullary spinal tumor, spinal ischemia, or polyneuropathy [3,11,20]. Early diagnosis is very important because this condition is a potentially treatable disease through surgical or embolization therapy. Obviously, embolization could not be an option for treatment of our patient. As a matter of fact, we prefer surgical obliteration in the spinal dural AVFs because surgical treatment is a safe, simple, and proven technique without major morbidity. Most recently, Atkinson et al [3] from Mayo Clinic reported their experience including 94 cases with dural AVF. They also stated that embolization is associated with an unacceptable high recurrence rate. The dural AVF in our patient is different from the dural AVFs reported previously because of venous aneurysm. Focal venous aneurysms are not infrequent in spinal cord arteriovenous malformations. In contrast, venous aneurysm with a spinal dural AVF has rarely been reported [6]. Theoretically, their association is an unexpected condition because dural AVFs are low-flow lesions [9]. This association cannot be explained as congenital vessels anomaly or merely coincidental. Although spinal dural AVFs are low-flow lesions, the development of venous aneurysm must be because of hemodynamic factors. In 1989, Hassler et al [7] investigated local hemodynamics of spinal dural AVFs by micro-
Spinal AVF and Venous Aneurysm
vascular Doppler sonography during the operations. According to their results, the mean venous pressure in the spinal dural AVFs is approximately 74% of the systemic arterial pressure. The occurrence of the venous aneurysms is probably because of this high venous pressure. In our case, it was difficult to identify clearly which pathology caused the symptoms. Venous hypertension may have played a major role in the pathogenesis of myelopathy, but the mass effect of the partially thrombosed aneurysm, 1 cm in size, also may have contributed to the patient’s symptoms. For this reason, it was excised.
Conclusions Our case illustrates two important points related to patients harboring spinal dural AVF. First, clinical suspicion and MR findings are important in diagnosis of these lesions, and surgical exploration is indicated despite negative or nondiagnostic angiographic results in such cases. Second, venous aneurysms may be associated with spinal dural AVFs, and although spinal dural AVFs are low-flow lesions, the development of the venous aneurysm is probably a result of high venous pressure
Surg Neurol 337 2003;60:334 – 8
11. 12.
13.
14. 15. 16.
17.
18. 19.
REFERENCES 1. Alleyne CH, Barrow DL, Joseph G. Surgical management of angiographically occult spinal dural arteriovenous fistula (type I spinal arteriovenous malformations): three technical case reports. Neurosurgery 1999;44:891–5. 2. Aminoff MJ, Barnard RO, Logue V. The pathophysiology of spinal vascular malformation. J Neurol Sci 1974;23:255–63. 3. Atkinson JLD, Miller GM, Krauss DE, et al. Clinical and radiographic features of dural arteriovenous fistula, a treatable cause of myelopathy. Mayo Clin Proc 2001; 76:1120 –30. 4. Bowen BC, Fraser K, Kochan JP, Pattany PM, Quencer RM. Spinal dural arteriovenous fistulas: evaluation with MR angiography. AJNR 1995;16:2029 –43. 5. Gilbertson JR, Miller GM, Goldman MS, Marsh WR. Spinal dural arteriovenous fistulas: MR and myelographic findings. AJNR 1995;16:2049 –57. 6. Glasser R, Mason R, Mickle JP, Peters KR. Embolization of dural arteriovenous fistula of the ventral cervical spinal canal in a nine-year old boy. Neurosurgery 1993;33:1089 –94. 7. Hassler W, Thorn A, Grote EH. Hemodynamics of spinal dural arteriovenous fistulas. An intraoperative study. J Neurosurg 1989;70:360 –70. 8. Herdt JR, DiChiro G, Doppman JL. Combined arterial and arteriovenous aneurysms of the spinal cord. Radiology 1971;99:589 –93. 9. Hurst RW, Kenyon RC, Lavi E, Raps EC, Marcotte P. Spinal dural arteriovenous fistula. The pathology of venous hypertensive myelopathy. Neurology 1995;45: 1309 –13. 10. Kendall BE, Logue V. Spinal epidural angiomatous
20.
malformations draining to intrathecal veins. Neuroradiology 1977;13:181–9. Lee TT, Gromelski EB, Bowen BC, Green BA. Diagnostic and surgical management of spinal dural arteriovenous fistulas. Neurosurgery 1998;43:242–7. Mascalchi M, Scazzeri F, Prosetti D, Ferrito G, Salvi F, Quilici N. Dural arteriovenous fistula at the craniocervical junction with perimedullary drainage. AJNR 1996;17:1137–41. McCutcheon IE, Doppman JL, Oldfield EH. Microvascular anatomy of dural arteriovenous abnormalities of the spine. A microangiographic study. J Neurosurg 1996;84:215–20. Merland JJ, Riche MC, Chirai C. Intraspinal extramedullary arteriovenous fistula draining into the medullary veins. J Neuroradiology 1980;7:271–320. Miyamoto S, Kikuchi H, Karasawa J, Ikota T, Nagata I. Spinal cord arteriovenous malformations associated with spinal aneurysms. Neurosurgery 1983;13:577–80. Mourier KL, Gelbert F, Rey A, et al. Spinal dural arteriovenous malformations with perimedullary drainage. Indications and results of surgery in 30 cases. Acta Neurochir 1989;100:136 –41. Rosenblum B, Oldfield EH, Doppman JL, DiChiro G. Spinal arteriovenous malformations. A comparison of dural arteriovenous fistulas and intradural AVM’s in 81 patients. J Neurosurg 1987;67:795–802. Symon L, Kuyama H, Kendall B. Dural arteriovenous malformations of the spine. Clinical features and surgical results in 55 cases. J Neurosurg 1994;60:238 –47. Terwey B, Becker H, Thorn AK, Vahldiek G. Gadolinium-DTPA enhanced MR imaging of spinal dural arteriovenous fistulas. J Comput Assist Tomogr 1989;13:30 –7. Westphal M, Koch C. Management of spinal dural arteriovenous fistulae using an interdisciplinary neuroradiological/neurosurgical approach: experience with 47 cases. Neurosurgery 1999;45:451–8.
COMMENTARY
This case exemplifies the unusual nature of spinal dural arteriovenous fistulae, and reinforces surgical exploration in textbook cases of clinical course and MR imaging despite nondiagnostic spinal angiography. In our previously published operated 94 cases of dural AVF [1], there were 4 cases of venous varix identified in high-flow lesions. However, all were localized to the fistula draining vein site, making this case report of parenchymal location distant from the fistula connection quite remarkable. All but 1 of our 94 patients improved in the postoperative period, but patients presenting in poor grade (5T or 5P modified Aminoff-Logue Scale) improve the least, as this case typifies. Bolus contrast MR of the spine with coronal imaging localizes the fistula most of the time. The image quality lacks the definitive localization of angiography, but may direct the angiogram to a more focal region of the spine. In cases such as this, where technical problems precluded diagnostic angiogram, a region (left or right, generalized segmen-
Spinal Dural Arteriovenous Fistula and Venous Aneurysm: An Unusual Association ¨ zgen, M.D. Servet Inci, M.D., Ph.D, Pınar Akdemir, M.D., and Tuncalp O Department of Neurosurgery, School of Medicine, University of Hacettepe, Ankara, Turkey
¨ zgen T. Spinal dural arteriovenous fistula and Inci S, Akdemir P, O venous aneurysm: an unusual association. Surg Neurol 2003;60: 334 – 8. OBJECTIVE
Although spinal intradural arteriovenous malformations have frequently been associated with venous aneurysms, spinal dural arteriovenous fistulas typically are not. We describe a case of conus medullaris compression by a large saccular venous aneurysm of a spinal dural arteriovenous fistula. As these lesions are slow-flow arteriovenous shunts, development of a venous aneurysm is an unexpected condition. Pathogenesis of these aneurysms is briefly discussed. METHODS
A 61-year-old man presented with progressive spastic paraparesis of 1-year duration. Spinal magnetic resonance (MR) suggested abnormal serpiginous vessels and a partially thrombosed aneurysm at the level of conus medullaris. Spinal angiography was performed but some lower thoracal and upper lumbar segments could not be catheterized because of severe atherosclerosis. As a result, vascular anomaly could not be demonstrated angiographically, but surgical exploration was found to be necessary because of positive clinical and MR findings. RESULTS
An arteriovenous fistula that entered the spinal canal beneath the L1 pedicle on the left was identified. It was easily coagulated and cut. The partially thrombosed aneurysm within the conus medullaris was also resected. CONCLUSION
Our case illustrates two important points related to patients harboring spinal dural arteriovenous fistulas (AVF): First, clinical suspicion and MR findings are important in diagnosis of these lesions and surgical exploration is indicated despite negative or nondiagnostic angiographic results in such cases. Second, venous aneurysms may be associated with spinal dural AVFs, and although spinal dural AVFs are low-flow lesions, the development of the venous aneurysm is probably a result of high venous pressure. © 2003 Elsevier Inc. All rights reserved. KEY WORDS
Spinal dural AVF, venous aneurysm, conus medullaris.
Address reprint requests to: Dr. Servet Inci, Emek Mahallesi, 4. Cadde, No:70/8, 06510 Ankara, Turkey. Received July 19, 2002; accepted December 30, 2003. 0090-3019/03/$–see front matter doi:10.1016/S0090-3019(03)00239-8
pinal dural arteriovenous fistulas (AVF) are abnormal communications between radicular arteries and perimedullary veins or coronal venous plexus [4,13]. The fistula is usually located within the dura covering of the spinal nerve root. These lesions, which are the most common type of spinal vascular malformations in adults, cause a gradually progressive myelopathy as a result of venous hypertension in the spinal cord. Although intradural spinal arteriovenous malformations are frequently associated with venous aneurysms [8,15], spinal dural AVFs are typically not associated with venous aneurysms in the large series [16-20]. We report a case of typical saccular venous aneurysm presenting with conus medullaris compression.
S
Case Reports A 61-year-old man presented with progressive spastic paraparesis of 1-year duration. The patient was evaluated in a local hospital and was referred to our department for diagnosis and treatment. Neurologic examination revealed severe weakness of both legs involving the proximal and distal muscles. He was not ambulatory and had urinary retention, decreased anal spinchter tone, and decreased sensation below T10 level. Spinal magnetic resonance (MR) showed multiple punctuate and serpiginous areas of low intensity on the dorsal surface of the spinal cord and conus medullaris. In addition, there was a rounded hypoisointense lesion at the level of conus on T2weighted sections, suggesting a thrombosed aneurysm (Figure 1). Another typical finding on the same image was increased signal intensity within the cord from T11 to the end of conus medullaris. A spinal vascular anomaly was considered. Spinal angiography was performed through the femoral route, but some lower thoracal and upper © 2003 Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010 –1710
Spinal AVF and Venous Aneurysm
Surg Neurol 335 2003;60:334 – 8
2
Histologic appearance of the partially thrombosed aneurysm.
lumbar segments, including the artery of Adamkiewicz, could not be catheterized because of severe atherosclerosis. So vascular anomaly could not be demonstrated angiographically, but surgical exploration was found to be necessary because of positive clinical and MR findings.
was extended to L1. An arteriovenous fistula that entered the spinal canal beneath the L1 pedicle on the left was identified. It coursed from the inner surface of the dura to communicate with the dilated, tortuous vein. The fistula was dissected away from nerve root and easily coagulated and cut. The draining vein gradually collapsed within several minutes. Just as the vein collapsed, a rounded, dark blue, cystic lesion was noticed on the conus medullaris. This lesion was just distal to the fistula and buried in the conus medullaris. The well-circumscribed, nut-sized lesion was separated from the spinal cord with gentle dissection. The dome of the lesion was incised, with escape of chocolate-colored fluid. It was a partially thrombosed aneurysm that was resected because of mass effect. It was histologically verified as a partially thrombosed aneurysm (Figure 2).
OPERATION The thoracolumbar region from T9 to L2, because some of these segments could not be visualized on preoperative angiography, was cleaned and draped. Initially, T11 and T12 laminectomies were carried out. The dura was opened longitudinally under the microscope, and a dilated, tortuous vein was encountered on the dorsal surface of the cord. A systematic intradural exploration of each radicular artery at nerve root level was bilaterally examined, but a fistula could not be found. The artery of Adamkiewicz was also identified at left T12 level as normal caliber and appearance. Then laminectomy
POSTOPERATIVE COURSE Despite some initial improvement, the patient had no long-term change in his neurologic deficits. His long-standing lower extremity spasticity did not resolve but has not progressed over the subsequent 6 months. Control spinal angiography was not performed because AVF could not be demonstrated preoperatively. Control spinal MR was performed on postoperative 10th day, and T2-weighted sections showed no tortuous vessels or aneurysm. Furthermore, increased signal intensity within the spinal cord had been partially decreased (Figure 3).
Sagittal T2-weighted spinal MR demonstrates serpiginous signal void areas on the dorsal surface of the cord, increased signal intensity within the spinal cord from T11 to L2 and an intramedullary rounded mass (arrow) in the conus medullaris. At surgery, the mass was found to be a partially thrombosed venous aneurysm distal to the fistula.
1
336 Surg Neurol 2003;60:334 – 8
Postoperative sagittal T2-weighted spinal MR shows no tortuous vessels. Increased signal intensity within the spinal cord is mostly resolved.
3
Discussion Spinal DAVF was first described by Kendall and Louge [10] as “spinal epidural angiomatous malformations draining into intrathecal veins,” in 1977. Three years later, Merland et al [14] reported these lesions as “radiculomeningeal arteriovenous fistula.” Spinal dural AVF (also known as Type I arteriovenous malformation) is the most common spinal vascular malformation, representing more than two-thirds of all types [5,18]. They are usually found in the thoracolumbar region in middle-aged men [3,16,18], as in our case. Their pathogenesis is still debatable. However, they are considered to be acquired, probably as a result of venous thrombosis or trauma [13]. Such a fistula consists of a single arterial feeder (occasionally two or three feeders) from a radicular artery that enters the dura at the dural root sleeve. Spinal cord arteries never participate in the shunt. This is the most important difference of dural AVF from spinal cord AVF. The shunt is drained intradurally by retrograde flow through medullary vein, resulting in engorgement of the coronal venous plexus and intraparenchymal radial veins [17,18]. It is now well accepted that this venous hypertension accounts for the progressive
Inci et al
myelopathy. The arteriovenous fistula causes a chronic venous congestion of the spinal cord and secondary ischemia [2,9], which is responsible for the clinical picture of myelopathy. Symptoms are usually gradually progressive, consisting of weakness and sensory symptoms in the legs with a severe disturbance in micturition. Spinal angiography, of course, is the gold standard for the diagnosis of spinal dural AVFs. Our angiography was incomplete because of technical difficulties due to severe atherosclerosis. Rarely, arteriovenous fistulas cannot be demonstrated because of some other reasons: (1) The fistula itself can be microscopic [12]; (2) the shunt flow can be extremely slow [19]. Recently, Alleyne et al [1] also reported three patients with angiographically occult spinal dural AVFs. In these situations, a high index of clinical suspicion should be key in diagnosing these lesions, and surgical exploration is indicated despite negative angiographic results. Sometimes the location of the fistula may not correlate with the level and site of the clinical symptoms [17,18]. Therefore, the patient should be prepared for a large exploration. Spinal dural AVFs are frequently diagnosed after delays because they are usually not included in the differential diagnosis of myelopathy. Sometimes, they are also misdiagnosed as transverse myelitis, intramedullary spinal tumor, spinal ischemia, or polyneuropathy [3,11,20]. Early diagnosis is very important because this condition is a potentially treatable disease through surgical or embolization therapy. Obviously, embolization could not be an option for treatment of our patient. As a matter of fact, we prefer surgical obliteration in the spinal dural AVFs because surgical treatment is a safe, simple, and proven technique without major morbidity. Most recently, Atkinson et al [3] from Mayo Clinic reported their experience including 94 cases with dural AVF. They also stated that embolization is associated with an unacceptable high recurrence rate. The dural AVF in our patient is different from the dural AVFs reported previously because of venous aneurysm. Focal venous aneurysms are not infrequent in spinal cord arteriovenous malformations. In contrast, venous aneurysm with a spinal dural AVF has rarely been reported [6]. Theoretically, their association is an unexpected condition because dural AVFs are low-flow lesions [9]. This association cannot be explained as congenital vessels anomaly or merely coincidental. Although spinal dural AVFs are low-flow lesions, the development of venous aneurysm must be because of hemodynamic factors. In 1989, Hassler et al [7] investigated local hemodynamics of spinal dural AVFs by micro-
Spinal AVF and Venous Aneurysm
vascular Doppler sonography during the operations. According to their results, the mean venous pressure in the spinal dural AVFs is approximately 74% of the systemic arterial pressure. The occurrence of the venous aneurysms is probably because of this high venous pressure. In our case, it was difficult to identify clearly which pathology caused the symptoms. Venous hypertension may have played a major role in the pathogenesis of myelopathy, but the mass effect of the partially thrombosed aneurysm, 1 cm in size, also may have contributed to the patient’s symptoms. For this reason, it was excised.
Conclusions Our case illustrates two important points related to patients harboring spinal dural AVF. First, clinical suspicion and MR findings are important in diagnosis of these lesions, and surgical exploration is indicated despite negative or nondiagnostic angiographic results in such cases. Second, venous aneurysms may be associated with spinal dural AVFs, and although spinal dural AVFs are low-flow lesions, the development of the venous aneurysm is probably a result of high venous pressure
Surg Neurol 337 2003;60:334 – 8
11. 12.
13.
14. 15. 16.
17.
18. 19.
REFERENCES 1. Alleyne CH, Barrow DL, Joseph G. Surgical management of angiographically occult spinal dural arteriovenous fistula (type I spinal arteriovenous malformations): three technical case reports. Neurosurgery 1999;44:891–5. 2. Aminoff MJ, Barnard RO, Logue V. The pathophysiology of spinal vascular malformation. J Neurol Sci 1974;23:255–63. 3. Atkinson JLD, Miller GM, Krauss DE, et al. Clinical and radiographic features of dural arteriovenous fistula, a treatable cause of myelopathy. Mayo Clin Proc 2001; 76:1120 –30. 4. Bowen BC, Fraser K, Kochan JP, Pattany PM, Quencer RM. Spinal dural arteriovenous fistulas: evaluation with MR angiography. AJNR 1995;16:2029 –43. 5. Gilbertson JR, Miller GM, Goldman MS, Marsh WR. Spinal dural arteriovenous fistulas: MR and myelographic findings. AJNR 1995;16:2049 –57. 6. Glasser R, Mason R, Mickle JP, Peters KR. Embolization of dural arteriovenous fistula of the ventral cervical spinal canal in a nine-year old boy. Neurosurgery 1993;33:1089 –94. 7. Hassler W, Thorn A, Grote EH. Hemodynamics of spinal dural arteriovenous fistulas. An intraoperative study. J Neurosurg 1989;70:360 –70. 8. Herdt JR, DiChiro G, Doppman JL. Combined arterial and arteriovenous aneurysms of the spinal cord. Radiology 1971;99:589 –93. 9. Hurst RW, Kenyon RC, Lavi E, Raps EC, Marcotte P. Spinal dural arteriovenous fistula. The pathology of venous hypertensive myelopathy. Neurology 1995;45: 1309 –13. 10. Kendall BE, Logue V. Spinal epidural angiomatous
20.
malformations draining to intrathecal veins. Neuroradiology 1977;13:181–9. Lee TT, Gromelski EB, Bowen BC, Green BA. Diagnostic and surgical management of spinal dural arteriovenous fistulas. Neurosurgery 1998;43:242–7. Mascalchi M, Scazzeri F, Prosetti D, Ferrito G, Salvi F, Quilici N. Dural arteriovenous fistula at the craniocervical junction with perimedullary drainage. AJNR 1996;17:1137–41. McCutcheon IE, Doppman JL, Oldfield EH. Microvascular anatomy of dural arteriovenous abnormalities of the spine. A microangiographic study. J Neurosurg 1996;84:215–20. Merland JJ, Riche MC, Chirai C. Intraspinal extramedullary arteriovenous fistula draining into the medullary veins. J Neuroradiology 1980;7:271–320. Miyamoto S, Kikuchi H, Karasawa J, Ikota T, Nagata I. Spinal cord arteriovenous malformations associated with spinal aneurysms. Neurosurgery 1983;13:577–80. Mourier KL, Gelbert F, Rey A, et al. Spinal dural arteriovenous malformations with perimedullary drainage. Indications and results of surgery in 30 cases. Acta Neurochir 1989;100:136 –41. Rosenblum B, Oldfield EH, Doppman JL, DiChiro G. Spinal arteriovenous malformations. A comparison of dural arteriovenous fistulas and intradural AVM’s in 81 patients. J Neurosurg 1987;67:795–802. Symon L, Kuyama H, Kendall B. Dural arteriovenous malformations of the spine. Clinical features and surgical results in 55 cases. J Neurosurg 1994;60:238 –47. Terwey B, Becker H, Thorn AK, Vahldiek G. Gadolinium-DTPA enhanced MR imaging of spinal dural arteriovenous fistulas. J Comput Assist Tomogr 1989;13:30 –7. Westphal M, Koch C. Management of spinal dural arteriovenous fistulae using an interdisciplinary neuroradiological/neurosurgical approach: experience with 47 cases. Neurosurgery 1999;45:451–8.
COMMENTARY
This case exemplifies the unusual nature of spinal dural arteriovenous fistulae, and reinforces surgical exploration in textbook cases of clinical course and MR imaging despite nondiagnostic spinal angiography. In our previously published operated 94 cases of dural AVF [1], there were 4 cases of venous varix identified in high-flow lesions. However, all were localized to the fistula draining vein site, making this case report of parenchymal location distant from the fistula connection quite remarkable. All but 1 of our 94 patients improved in the postoperative period, but patients presenting in poor grade (5T or 5P modified Aminoff-Logue Scale) improve the least, as this case typifies. Bolus contrast MR of the spine with coronal imaging localizes the fistula most of the time. The image quality lacks the definitive localization of angiography, but may direct the angiogram to a more focal region of the spine. In cases such as this, where technical problems precluded diagnostic angiogram, a region (left or right, generalized segmen-