Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mascalchi, M.
Right arrow Articles by Villari, N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mascalchi, M.
Right arrow Articles by Villari, N.
(Radiology. 2001;219:346-353.)
© RSNA, 2001

Neuroradiology

Spinal Vascular Malformations: MR Angiography after Treatment1

Mario Mascalchi, MD, PhD, Giampiero Ferrito, MD, Nello Quilici, MD, Salvatore Mangiafico, MD, Mirco Cosottini, MD, Martino Cellerini, MD, Letterio S. Politi, MD, Laura Guerrini, MD, Carlo Bartolozzi, MD and Natale Villari, MD

1 From the Department of Clinical Pathophysiology, Section of Radiodiagnostics, University of Florence, Viale Morgagni 85, 50134 Florence, Italy (M.M., M. Cellerini, L.S.P., L.G., N.V.); Department of Neuroradiology, Hospital Riuniti, Livorno, Italy (G.F., N.Q.); Department of Neuroradiology, Hospital of Careggi, Florence, Italy (S.M.); and Department of Radiology, University of Pisa, Italy (M. Cosottini, C.B.). Received April 25, 2000; revision requested June 12; revision received July 28; accepted August 30. Address correspondence to M.M.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
PURPOSE: To evaluate the role of magnetic resonance (MR) angiography in the assessment of spinal vascular malformation therapy.

MATERIALS AND METHODS: Thirty-four patients with spinal vascular malformations (30 dural arteriovenous fistulas, two perimedullary arteriovenous fistulas, and two intramedullary arteriovenous malformations) underwent MR angiography and MR imaging before and after endovascular or surgical treatment.

RESULTS: MR angiography showed residual flow in perimedullary vessels in seven patients with dural fistula after embolization with liquid adhesive. In all seven, treatment failure was confirmed with arteriography. Long-lasting disappearance of flow in perimedullary vessels was demonstrated at MR angiography in 22 patients with dural fistula. MR imaging demonstrated normalization of spinal cord volume in 16 of 22 patients and signal intensity on T2-weighted images in three patients. Disappearance of cord enhancement was observed in five of 21 patients and of perimedullary enhanced vessels in six of 13 patients. In one additional patient with dural fistula treated with embolization, early posttreatment MR angiography showed disappearance of flow in perimedullary vessels, which reappeared at follow-up and was consistent with reopening of a small residual fistula. Posttreatment MR angiography demonstrated transient reduction of flow in the nidus in two patients with intramedullary malformations treated with embolization. Permanent disappearance of flow in the perimedullary vessel was seen after endovascular treatment in two patients with perimedullary fistula.

CONCLUSION: MR angiography is more sensitive than MR imaging in depicting residual or recurrent flow in peri- or intramedullary vessels, which indicates patency of the vascular malformation.

Index terms: Arteriovenous malformations, spinal, 30.149 • Arteriovenous malformations, therapeutic embolization, 30.1264 • Magnetic resonance (MR), vascular studies, 30.121411, 30.12142, 30.12143 • Spinal cord, abnormalities, 30.149 • Spinal cord, blood supply, 30.149 • Spinal cord, MR, 30.121411, 30.12142, 30.12143


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Vascular malformations are a treatable cause of myelopathy. Magnetic resonance (MR) angiography is a useful complement to MR imaging for detection and characterization of spinal vascular malformations before digital subtraction angiography (DSA) (1–6). To our knowledge, the role of MR angiography in the evaluation of endovascular or surgical treatment of these lesions has not been thoroughly investigated. The purpose of our study was to report posttreatment MR angiographic and MR imaging findings in 34 patients with spinal vascular malformations.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
From April 1992 to February 2000, 34 patients (10 women, 24 men; age range, 18–84 years) with native spinal vascular malformations underwent 122 spinal MR examinations before (n = 46) and after (n = 76) treatment in three centers (University of Florence; University of Pisa; Hospital Riuniti, Livorno, Italy) as a part of our routine evaluation. All MR examinations included MR angiographic and MR imaging acquisitions. Selective DSA was performed within 2 weeks after pretreatment MR examination; DSA diagnosis (7,8) included spinal dural arteriovenous fistula (n = 28) and dural arteriovenous fistula at the cranial-cervical junction (n = 2) with perimedullary venous drainage in 30 patients, as well as intradural arteriovenous malformations in four patients. The latter comprised two cases of intramedullary arteriovenous malformations of a glomus type and one case each of a low-flow (type 1) and a high-flow (type 3) perimedullary arteriovenous fistula. The level and side of the dural arteriovenous fistulas are reported in the Table. A cervical arteriovenous malformation of the glomus type was fed by the left cervical ascending artery, and a thoracic arteriovenous malformation of the glomus type was fed by the left T10 intercostal artery. The type 1 perimedullary arteriovenous fistula had a single feeder vessel arising from the left L1 lumbar artery. Two afferent vessels arising from the right L1 and L4 lumbar arteries fed the type 3 perimedullary arteriovenous fistula.


View this table:
[in this window]
[in a new window]

  		Level and Side of the Dural Arteriovenous Fistula with Perimedullary Venous Drainage in 30 Patients

 
Nineteen patients with intracranial (n = 1) or spinal (n = 18) dural arteriovenous fistula underwent embolization with endovascular liquid adhesive (N-butyl cyanoacrylate), whereas 11 patients with intracranial (n = 1) or spinal (n = 10) dural arteriovenous fistula underwent surgery. The mean interval between treatment and first posttreatment MR examination in the patients with dural arteriovenous fistula was 54 days (range, 1–95 days). In 13 of them, an early (ie, within 1 week) posttreatment follow-up examination was performed. The mean interval between treatment and the last follow-up posttreatment MR examination in the same group of patients was 8 months (range, 1–42 months).

The patient with cervical intramedullary arteriovenous malformation was examined 1 and 14 months after embolization with particles. The patient with thoracic intramedullary arteriovenous malformation was examined 6, 21, 32, 41, and 76 months after embolization with liquid adhesive. The patient with type 1 perimedullary arteriovenous fistula was examined 5, 13, and 24 months after embolization with liquid adhesive. The patient with type 3 perimedullary arteriovenous fistula underwent MR follow-up examinations 12, 31, and 44 months after embolization with liquid adhesive and one Guglielmi detachable coil.

Clinical outcome of treatment was based on the results of the neurologic examination and on the subjective evaluation by the patient at the time of posttreatment MR examination.

Posttreatment DSA was performed in 10 patients (eight with dural arteriovenous fistula and two with intramedullary arteriovenous malformation) on the basis of the evidence of persistent flow in abnormal intrathecal vessels at MR angiography. In only two patients (one with dural arteriovenous fistula and one with type 3 perimedullary arteriovenous fistula), posttreatment DSA was performed notwithstanding the disappearance of abnormal vessels at posttreatment MR angiography.

MR Protocols
MR angiography.—According to the evolution of systems hardware and software, three MR angiographic protocols were used during the study.

First protocol (April 1992 through September 1994): Seventy-four (27 pretreatment and 47 posttreatment) MR angiographic examinations were performed with a 0.5-T system (MR Max Plus; GE Medical Systems, Milwaukee, Wis) equipped with 7.7-mT/m gradients. A quadrature surface coil was used for examinations of the thoracolumbar spine, and a linear surface coil was used for the cervical spine. Two-dimensional (2D) phase-contrast MR angiography (60–70/27 [repetition time msec/echo time msec]; flip angle, 30°; 16 or 32 signals acquired) was performed after manual intravenous administration of gadopentetate dimeglumine (0.1–0.3 mmol per kilogram of body weight). Sagittal and coronal single 15-mm-thick sections (field of view, 25–30 cm x 25–30 cm; imaging matrix, 160 x 192–224) were acquired, with bipolar gradients along the three body axes, which were set to optimally demonstrate flow of 6 cm/sec in the case of dural arteriovenous fistula and 20–30 cm/sec in the case of intradural arteriovenous malformations. Acquisition and reconstruction of each sagittal and coronal MR angiogram required 15 minutes.

Second protocol (October 1994 through February 1998): Thirty (14 pretreatment and 16 posttreatment) MR angiographic examinations were performed with a 1.5-T system (Signa, GE Medical Systems) equipped with 12-mT/m gradients. A quadrature surface coil was used for the thoracolumbar spine, and a quadrature volume neck coil was used for the cervical spine. In addition to 2D phase-contrast MR angiographic (38/11; flip angle, 20°–30°; eight or 12 signals acquired) single-section (field of view, 25–30 cm x 25–30 cm; imaging matrix, 192 x 256) acquisitions in the sagittal plane, volume three-dimensional (3D) phase-contrast MR angiography (29–33/9–11; flip angle, 20°–30°; one or two signals acquired) was performed, with coronal acquisition of 28–40 partitions (field of view, 25–30 cm x 25–30 cm; imaging matrix, 192 x 256) 1–2 mm thick. The 3D data set was viewed along different projections with the maximum intensity projection algorithm. Characteristics of bipolar gradients and injection of contrast material were as in the first protocol. Acquisition time for 2D MR angiography was less than 5 minutes and for 3D phase-contrast MR angiography was about 10 minutes. Reconstruction of the 3D data set required a few minutes.

Third protocol (March 1998 through February 2000): Eighteen (five pretreatment and 13 posttreatment) MR angiographic examinations were performed with two 1.5-T systems (Signa, GE Medical Systems; Gyroscan NT, Philips Medical Systems, Best, The Netherlands) equipped with high-performance (23 mT/m) gradients and quadrature or phased-array surface coils. Contrast-enhanced time-resolved MR angiography was performed with a fast 3D sequence (6.0–6.9/1.2–1.8; flip angle, 40°; one signal acquired), which enabled acquisition of 20–32 partitions (field of view, 25 cm x 25–35 cm; imaging matrix, 160 or 256 x 256 or 512) 1.5–2.0 mm thick in 17–20 seconds in the coronal plane. The sequence was run four to five times with a 1-second delay between acquisitions, and its start was synchronous with the administration of 0.1–0.3 mmol/kg gadopentetate dimeglumine or gadodiamide in an antecubital vein with an automatic power injector (velocity of flow, 3 mL/sec). No breath holding was requested of the patient. Contrast-enhanced time-resolved MR angiography was followed by 2D or 3D phase-contrast MR angiography as in the second protocol. Acquisition time for contrast-enhanced MR angiography was 3 minutes, but reconstruction of the serial 3D data sets required about 10 minutes. The 3D contrast-enhanced time-resolved and 3D phase-contrast MR angiographic data were viewed along different projections with the maximum intensity projection algorithm.

MR imaging.—MR angiography was always preceded by sagittal T1-weighted (360–500/20–30 with two or four signals acquired) spin-echo and T2-weighted (1,800–5,000/90–100 with one, two, or four signals acquired) spin-echo or turbo spin-echo sequences (field of view, 25–35 cm x 25–35 cm; imaging matrix, 256 x 256; section thickness, 3–5 mm with 0.3–1.0 gap). Contrast-enhanced sagittal T1-weighted (360–500/20–30 with two or four signals acquired) spin-echo sequences were used after MR angiography in all but six patients (two with dural arteriovenous fistulas and four with intramedullary arteriovenous malformations) for the pretreatment examinations, and in all patients for the posttreatment examination. Because of intervening acquisition and reconstruction of MR angiograms lasting several minutes, T1-weighted sequences used after the administration of contrast material produced delayed contrast-enhanced images.

Data Analysis
After completion of the study in February 2000, two observers (M.M., M. Cosottini) jointly performed an unblinded review of the hard copies of the DSA studies and of all pre- and posttreatment MR angiographic and MR imaging examinations.

In patients with intracranial or spinal dural arteriovenous fistula with perimedullary venous drainage, MR angiographic examinations were evaluated only for demonstration and extension of flow in perimedullary vessels without any attempt to identify the level and side of the dural fistula. The MR imaging features considered included cord swelling, presence and extent of cord hyperintensity in T2-weighted images, cord contrast enhancement, and evidence and extent of perimedullary vessels as serpentine or punctate areas of signal void in T2-weighted images or of contrast enhancement in postcontrast T1-weighted images.

In patients with intradural arteriovenous malformations, flow in intramedullary and perimedullary vessels was evaluated at MR angiography before and after treatment, and results were compared with evidence of signal void in the same areas in nonenhanced T1- and T2-weighted MR images. Moreover, signal intensity changes and contrast enhancement of the spinal cord were noted.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Dural Arteriovenous Fistulas before Treatment
MR angiography.—Flow in serpentine perimedullary structures corresponding to dilated vessels was demonstrated at MR angiography in all 30 patients (100%) (Figs 1, 2). Extension of these vessels ranged between two and 19 vertebral segments (mean, 6.5) and always included the level corresponding to the source of the fistula identified at DSA.



View larger version (119K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a. Spinal dural arteriovenous fistula (a) before and (b) after endovascular therapy. (a) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) obtained at patient presentation demonstrates flow in serpentine antero- and retromedullary vessels (arrows). (b) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) obtained the day after embolization of the fistula with liquid adhesive reveals persistent flow in anteromedullary vessels (solid arrow), which is consistent with treatment failure and which was confirmed at DSA. The ill-defined areas of faint flow signal intensity in the posterior portion of the thoracic spinal canal that correspond to cerebrospinal fluid flow phenomena (open arrow) normally can be observed at spinal phase-contrast MR angiography when low-velocity flow-encoding gradients are used.

 


View larger version (143K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b. Spinal dural arteriovenous fistula (a) before and (b) after endovascular therapy. (a) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) obtained at patient presentation demonstrates flow in serpentine antero- and retromedullary vessels (arrows). (b) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) obtained the day after embolization of the fistula with liquid adhesive reveals persistent flow in anteromedullary vessels (solid arrow), which is consistent with treatment failure and which was confirmed at DSA. The ill-defined areas of faint flow signal intensity in the posterior portion of the thoracic spinal canal that correspond to cerebrospinal fluid flow phenomena (open arrow) normally can be observed at spinal phase-contrast MR angiography when low-velocity flow-encoding gradients are used.

 


View larger version (143K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a. Spinal dural arteriovenous fistula (a-c) before and (d-h) after endovascular therapy. (a) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained at patient presentation shows enlargement of the cord at T11 and enhanced posterior perimedullary vessels (solid arrows) at T10 to T11. Punctate areas of signal void indenting the anterior cord surface are appreciable at T9 (open arrow). (b) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) demonstrates flow in serpentine posterior perimedullary vessels at T8 to T11 and T6 to T7 (arrow). (c) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery demonstrates a dural arteriovenous fistula (arrow) draining into the perimedullary venous plexus through a short tortuous intradural vein. (d) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained the day after embolization shows enhanced retromedullary vessels (arrow) at T11. (e) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) shows that the vessels do not exhibit significant flow. The linear areas of faint flow signal intensity in the posterior portion of the thoracic spinal canal correspond to cerebrospinal fluid flow phenomena (arrows). Treatment yielded clinical improvement for 2 months and was followed by progressive worsening. (f) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained 5 months after therapy shows enhanced serpentine vessels (arrows) anterior to the cord at T6 to T11 and posterior to the cord at T6 and T10 to T11 (arrows). (g) Coronal 3D phase-contrast MR angiogram (33/9; flip angle, 20°; two signals acquired) demonstrates flow in the perimedullary vessels (arrow) at T11 to T12. (h) Anteroposterior DSA image obtained after selective catheterization of the left T9 intercostal artery demonstrates that the fistula recruited small branches (solid arrows) from the left T9 intercostal artery, which is consistent with the reopening of a small residual fistula, and drains into the enlarged perimedullary venous plexus (open arrow).

 


View larger version (122K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b. Spinal dural arteriovenous fistula (a-c) before and (d-h) after endovascular therapy. (a) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained at patient presentation shows enlargement of the cord at T11 and enhanced posterior perimedullary vessels (solid arrows) at T10 to T11. Punctate areas of signal void indenting the anterior cord surface are appreciable at T9 (open arrow). (b) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) demonstrates flow in serpentine posterior perimedullary vessels at T8 to T11 and T6 to T7 (arrow). (c) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery demonstrates a dural arteriovenous fistula (arrow) draining into the perimedullary venous plexus through a short tortuous intradural vein. (d) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained the day after embolization shows enhanced retromedullary vessels (arrow) at T11. (e) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) shows that the vessels do not exhibit significant flow. The linear areas of faint flow signal intensity in the posterior portion of the thoracic spinal canal correspond to cerebrospinal fluid flow phenomena (arrows). Treatment yielded clinical improvement for 2 months and was followed by progressive worsening. (f) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained 5 months after therapy shows enhanced serpentine vessels (arrows) anterior to the cord at T6 to T11 and posterior to the cord at T6 and T10 to T11 (arrows). (g) Coronal 3D phase-contrast MR angiogram (33/9; flip angle, 20°; two signals acquired) demonstrates flow in the perimedullary vessels (arrow) at T11 to T12. (h) Anteroposterior DSA image obtained after selective catheterization of the left T9 intercostal artery demonstrates that the fistula recruited small branches (solid arrows) from the left T9 intercostal artery, which is consistent with the reopening of a small residual fistula, and drains into the enlarged perimedullary venous plexus (open arrow).

 


View larger version (162K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c. Spinal dural arteriovenous fistula (a-c) before and (d-h) after endovascular therapy. (a) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained at patient presentation shows enlargement of the cord at T11 and enhanced posterior perimedullary vessels (solid arrows) at T10 to T11. Punctate areas of signal void indenting the anterior cord surface are appreciable at T9 (open arrow). (b) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) demonstrates flow in serpentine posterior perimedullary vessels at T8 to T11 and T6 to T7 (arrow). (c) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery demonstrates a dural arteriovenous fistula (arrow) draining into the perimedullary venous plexus through a short tortuous intradural vein. (d) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained the day after embolization shows enhanced retromedullary vessels (arrow) at T11. (e) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) shows that the vessels do not exhibit significant flow. The linear areas of faint flow signal intensity in the posterior portion of the thoracic spinal canal correspond to cerebrospinal fluid flow phenomena (arrows). Treatment yielded clinical improvement for 2 months and was followed by progressive worsening. (f) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained 5 months after therapy shows enhanced serpentine vessels (arrows) anterior to the cord at T6 to T11 and posterior to the cord at T6 and T10 to T11 (arrows). (g) Coronal 3D phase-contrast MR angiogram (33/9; flip angle, 20°; two signals acquired) demonstrates flow in the perimedullary vessels (arrow) at T11 to T12. (h) Anteroposterior DSA image obtained after selective catheterization of the left T9 intercostal artery demonstrates that the fistula recruited small branches (solid arrows) from the left T9 intercostal artery, which is consistent with the reopening of a small residual fistula, and drains into the enlarged perimedullary venous plexus (open arrow).

 


View larger version (123K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2d. Spinal dural arteriovenous fistula (a-c) before and (d-h) after endovascular therapy. (a) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained at patient presentation shows enlargement of the cord at T11 and enhanced posterior perimedullary vessels (solid arrows) at T10 to T11. Punctate areas of signal void indenting the anterior cord surface are appreciable at T9 (open arrow). (b) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) demonstrates flow in serpentine posterior perimedullary vessels at T8 to T11 and T6 to T7 (arrow). (c) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery demonstrates a dural arteriovenous fistula (arrow) draining into the perimedullary venous plexus through a short tortuous intradural vein. (d) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained the day after embolization shows enhanced retromedullary vessels (arrow) at T11. (e) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) shows that the vessels do not exhibit significant flow. The linear areas of faint flow signal intensity in the posterior portion of the thoracic spinal canal correspond to cerebrospinal fluid flow phenomena (arrows). Treatment yielded clinical improvement for 2 months and was followed by progressive worsening. (f) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained 5 months after therapy shows enhanced serpentine vessels (arrows) anterior to the cord at T6 to T11 and posterior to the cord at T6 and T10 to T11 (arrows). (g) Coronal 3D phase-contrast MR angiogram (33/9; flip angle, 20°; two signals acquired) demonstrates flow in the perimedullary vessels (arrow) at T11 to T12. (h) Anteroposterior DSA image obtained after selective catheterization of the left T9 intercostal artery demonstrates that the fistula recruited small branches (solid arrows) from the left T9 intercostal artery, which is consistent with the reopening of a small residual fistula, and drains into the enlarged perimedullary venous plexus (open arrow).

 


View larger version (157K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2e. Spinal dural arteriovenous fistula (a-c) before and (d-h) after endovascular therapy. (a) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained at patient presentation shows enlargement of the cord at T11 and enhanced posterior perimedullary vessels (solid arrows) at T10 to T11. Punctate areas of signal void indenting the anterior cord surface are appreciable at T9 (open arrow). (b) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) demonstrates flow in serpentine posterior perimedullary vessels at T8 to T11 and T6 to T7 (arrow). (c) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery demonstrates a dural arteriovenous fistula (arrow) draining into the perimedullary venous plexus through a short tortuous intradural vein. (d) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained the day after embolization shows enhanced retromedullary vessels (arrow) at T11. (e) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) shows that the vessels do not exhibit significant flow. The linear areas of faint flow signal intensity in the posterior portion of the thoracic spinal canal correspond to cerebrospinal fluid flow phenomena (arrows). Treatment yielded clinical improvement for 2 months and was followed by progressive worsening. (f) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained 5 months after therapy shows enhanced serpentine vessels (arrows) anterior to the cord at T6 to T11 and posterior to the cord at T6 and T10 to T11 (arrows). (g) Coronal 3D phase-contrast MR angiogram (33/9; flip angle, 20°; two signals acquired) demonstrates flow in the perimedullary vessels (arrow) at T11 to T12. (h) Anteroposterior DSA image obtained after selective catheterization of the left T9 intercostal artery demonstrates that the fistula recruited small branches (solid arrows) from the left T9 intercostal artery, which is consistent with the reopening of a small residual fistula, and drains into the enlarged perimedullary venous plexus (open arrow).

 


View larger version (140K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2f. Spinal dural arteriovenous fistula (a-c) before and (d-h) after endovascular therapy. (a) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained at patient presentation shows enlargement of the cord at T11 and enhanced posterior perimedullary vessels (solid arrows) at T10 to T11. Punctate areas of signal void indenting the anterior cord surface are appreciable at T9 (open arrow). (b) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) demonstrates flow in serpentine posterior perimedullary vessels at T8 to T11 and T6 to T7 (arrow). (c) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery demonstrates a dural arteriovenous fistula (arrow) draining into the perimedullary venous plexus through a short tortuous intradural vein. (d) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained the day after embolization shows enhanced retromedullary vessels (arrow) at T11. (e) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) shows that the vessels do not exhibit significant flow. The linear areas of faint flow signal intensity in the posterior portion of the thoracic spinal canal correspond to cerebrospinal fluid flow phenomena (arrows). Treatment yielded clinical improvement for 2 months and was followed by progressive worsening. (f) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained 5 months after therapy shows enhanced serpentine vessels (arrows) anterior to the cord at T6 to T11 and posterior to the cord at T6 and T10 to T11 (arrows). (g) Coronal 3D phase-contrast MR angiogram (33/9; flip angle, 20°; two signals acquired) demonstrates flow in the perimedullary vessels (arrow) at T11 to T12. (h) Anteroposterior DSA image obtained after selective catheterization of the left T9 intercostal artery demonstrates that the fistula recruited small branches (solid arrows) from the left T9 intercostal artery, which is consistent with the reopening of a small residual fistula, and drains into the enlarged perimedullary venous plexus (open arrow).

 


View larger version (168K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2g. Spinal dural arteriovenous fistula (a-c) before and (d-h) after endovascular therapy. (a) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained at patient presentation shows enlargement of the cord at T11 and enhanced posterior perimedullary vessels (solid arrows) at T10 to T11. Punctate areas of signal void indenting the anterior cord surface are appreciable at T9 (open arrow). (b) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) demonstrates flow in serpentine posterior perimedullary vessels at T8 to T11 and T6 to T7 (arrow). (c) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery demonstrates a dural arteriovenous fistula (arrow) draining into the perimedullary venous plexus through a short tortuous intradural vein. (d) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained the day after embolization shows enhanced retromedullary vessels (arrow) at T11. (e) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) shows that the vessels do not exhibit significant flow. The linear areas of faint flow signal intensity in the posterior portion of the thoracic spinal canal correspond to cerebrospinal fluid flow phenomena (arrows). Treatment yielded clinical improvement for 2 months and was followed by progressive worsening. (f) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained 5 months after therapy shows enhanced serpentine vessels (arrows) anterior to the cord at T6 to T11 and posterior to the cord at T6 and T10 to T11 (arrows). (g) Coronal 3D phase-contrast MR angiogram (33/9; flip angle, 20°; two signals acquired) demonstrates flow in the perimedullary vessels (arrow) at T11 to T12. (h) Anteroposterior DSA image obtained after selective catheterization of the left T9 intercostal artery demonstrates that the fistula recruited small branches (solid arrows) from the left T9 intercostal artery, which is consistent with the reopening of a small residual fistula, and drains into the enlarged perimedullary venous plexus (open arrow).

 


View larger version (192K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2h. Spinal dural arteriovenous fistula (a-c) before and (d-h) after endovascular therapy. (a) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained at patient presentation shows enlargement of the cord at T11 and enhanced posterior perimedullary vessels (solid arrows) at T10 to T11. Punctate areas of signal void indenting the anterior cord surface are appreciable at T9 (open arrow). (b) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) demonstrates flow in serpentine posterior perimedullary vessels at T8 to T11 and T6 to T7 (arrow). (c) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery demonstrates a dural arteriovenous fistula (arrow) draining into the perimedullary venous plexus through a short tortuous intradural vein. (d) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained the day after embolization shows enhanced retromedullary vessels (arrow) at T11. (e) Sagittal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) shows that the vessels do not exhibit significant flow. The linear areas of faint flow signal intensity in the posterior portion of the thoracic spinal canal correspond to cerebrospinal fluid flow phenomena (arrows). Treatment yielded clinical improvement for 2 months and was followed by progressive worsening. (f) Sagittal postcontrast T1-weighted spin-echo MR image (400/20; four signals acquired) obtained 5 months after therapy shows enhanced serpentine vessels (arrows) anterior to the cord at T6 to T11 and posterior to the cord at T6 and T10 to T11 (arrows). (g) Coronal 3D phase-contrast MR angiogram (33/9; flip angle, 20°; two signals acquired) demonstrates flow in the perimedullary vessels (arrow) at T11 to T12. (h) Anteroposterior DSA image obtained after selective catheterization of the left T9 intercostal artery demonstrates that the fistula recruited small branches (solid arrows) from the left T9 intercostal artery, which is consistent with the reopening of a small residual fistula, and drains into the enlarged perimedullary venous plexus (open arrow).

 
MR imaging.—The spinal cord was enlarged in 24 patients (80%) (Fig 2) and of normal size in six. Signal hyperintensity in T2-weighted images was always present (100%), had variable extension (mean, 5.6 segments; range, 2–10), and had no consistent correlation with the level of the fistula. In particular, it was separated from the level of the fistula in 10 patients, as far as 17 segments in a case of intracranial dural fistula with perimedullary venous drainage.

Spinal cord contrast enhancement was present (11 intense and homogeneous; four intense and inhomogeneous; 10 mild and homogeneous) in 25 (89%) of the 28 patients who received contrast material and was absent in three (Fig 2). Extension of the area of cord enhancement substantially matched that of areas of increased signal intensity in T2-weighted images. Perimedullary vessels were seen as serpentine or punctate areas of signal void in 21 (70%) of 30 patients and as areas of contrast enhancement in 19 (67%) of 28 patients (Fig 2). In nine (30%) of 30 patients, no perimedullary vessels were seen in T2- or postcontrast T1-weighted images. Mean extension of the vessels was 2.6 segments (range, 1–7) in T2-weighted images, and 2.3 segments (range, 1–7) in postcontrast T1-weighted images. Extension of perimedullary vessels in T2-weighted images did not include the fistula level in nine patients. Also, extension of the contrast-enhanced vessel did not include the fistula level in eight patients.

Intradural Arteriovenous Malformations before Treatment
MR angiography.—Intrathecal serpentine structures exhibiting significant flow and corresponding mainly to the draining veins of the malformations were seen in all patients with intradural arteriovenous malformations (Fig 3). The nidus was depicted in two of the patients with intramedullary arteriovenous malformations (Fig 3). Dilatation of and increased flow in the intercostal arteries feeding the vascular malformations were observed in three patients (Fig 3).



View larger version (127K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a. Intramedullary arteriovenous malformation of the glomus type (a-c) before and (d, e) after endovascular therapy. (a) Sagittal nonenhanced T1-weighted spin-echo MR image (360/20; four signals acquired) obtained at patient presentation shows deformity of the spinal cord at T10 to T11 and multiple serpentine and punctate areas of signal void within (solid arrow) and at the anterior and posterior surfaces (open arrows) of the cord at the same level. (b) Coronal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) demonstrates an enlarged feeding vessel (solid arrow) originating from the left T10 intercostal artery, nidus (open arrow), and descending draining veins. (c) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery demonstrates the feeding artery (solid arrow) and the nidus (open arrow). (d) Coronal contrast-enhanced time-resolved MR angiogram (6/1.5; flip angle, 40°; one signal acquired) obtained 76 months after treatment demonstrates flow persistence in the nidus (open arrow) and predominant venous drainage through a left paramedian descending vein (solid arrow). (e) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery is used to confirm flow in the nidus (open arrow) and the venous drainage through a left paramedian descending vein (solid arrow).

 


View larger version (128K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b. Intramedullary arteriovenous malformation of the glomus type (a-c) before and (d, e) after endovascular therapy. (a) Sagittal nonenhanced T1-weighted spin-echo MR image (360/20; four signals acquired) obtained at patient presentation shows deformity of the spinal cord at T10 to T11 and multiple serpentine and punctate areas of signal void within (solid arrow) and at the anterior and posterior surfaces (open arrows) of the cord at the same level. (b) Coronal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) demonstrates an enlarged feeding vessel (solid arrow) originating from the left T10 intercostal artery, nidus (open arrow), and descending draining veins. (c) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery demonstrates the feeding artery (solid arrow) and the nidus (open arrow). (d) Coronal contrast-enhanced time-resolved MR angiogram (6/1.5; flip angle, 40°; one signal acquired) obtained 76 months after treatment demonstrates flow persistence in the nidus (open arrow) and predominant venous drainage through a left paramedian descending vein (solid arrow). (e) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery is used to confirm flow in the nidus (open arrow) and the venous drainage through a left paramedian descending vein (solid arrow).

 


View larger version (112K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3c. Intramedullary arteriovenous malformation of the glomus type (a-c) before and (d, e) after endovascular therapy. (a) Sagittal nonenhanced T1-weighted spin-echo MR image (360/20; four signals acquired) obtained at patient presentation shows deformity of the spinal cord at T10 to T11 and multiple serpentine and punctate areas of signal void within (solid arrow) and at the anterior and posterior surfaces (open arrows) of the cord at the same level. (b) Coronal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) demonstrates an enlarged feeding vessel (solid arrow) originating from the left T10 intercostal artery, nidus (open arrow), and descending draining veins. (c) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery demonstrates the feeding artery (solid arrow) and the nidus (open arrow). (d) Coronal contrast-enhanced time-resolved MR angiogram (6/1.5; flip angle, 40°; one signal acquired) obtained 76 months after treatment demonstrates flow persistence in the nidus (open arrow) and predominant venous drainage through a left paramedian descending vein (solid arrow). (e) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery is used to confirm flow in the nidus (open arrow) and the venous drainage through a left paramedian descending vein (solid arrow).

 


View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3d. Intramedullary arteriovenous malformation of the glomus type (a-c) before and (d, e) after endovascular therapy. (a) Sagittal nonenhanced T1-weighted spin-echo MR image (360/20; four signals acquired) obtained at patient presentation shows deformity of the spinal cord at T10 to T11 and multiple serpentine and punctate areas of signal void within (solid arrow) and at the anterior and posterior surfaces (open arrows) of the cord at the same level. (b) Coronal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) demonstrates an enlarged feeding vessel (solid arrow) originating from the left T10 intercostal artery, nidus (open arrow), and descending draining veins. (c) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery demonstrates the feeding artery (solid arrow) and the nidus (open arrow). (d) Coronal contrast-enhanced time-resolved MR angiogram (6/1.5; flip angle, 40°; one signal acquired) obtained 76 months after treatment demonstrates flow persistence in the nidus (open arrow) and predominant venous drainage through a left paramedian descending vein (solid arrow). (e) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery is used to confirm flow in the nidus (open arrow) and the venous drainage through a left paramedian descending vein (solid arrow).

 


View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3e. Intramedullary arteriovenous malformation of the glomus type (a-c) before and (d, e) after endovascular therapy. (a) Sagittal nonenhanced T1-weighted spin-echo MR image (360/20; four signals acquired) obtained at patient presentation shows deformity of the spinal cord at T10 to T11 and multiple serpentine and punctate areas of signal void within (solid arrow) and at the anterior and posterior surfaces (open arrows) of the cord at the same level. (b) Coronal 2D phase-contrast MR angiogram (60/27; flip angle, 30°; 12 signals acquired) demonstrates an enlarged feeding vessel (solid arrow) originating from the left T10 intercostal artery, nidus (open arrow), and descending draining veins. (c) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery demonstrates the feeding artery (solid arrow) and the nidus (open arrow). (d) Coronal contrast-enhanced time-resolved MR angiogram (6/1.5; flip angle, 40°; one signal acquired) obtained 76 months after treatment demonstrates flow persistence in the nidus (open arrow) and predominant venous drainage through a left paramedian descending vein (solid arrow). (e) Anteroposterior DSA image obtained after selective catheterization of the left T10 intercostal artery is used to confirm flow in the nidus (open arrow) and the venous drainage through a left paramedian descending vein (solid arrow).

 
MR imaging.—Serpentine or punctate areas of signal void were observed within or at the surface of the spinal cord in three patients (Fig 3). In the patient with type 1 perimedullary arteriovenous fistula, there was no evidence of abnormal peri- or intramedullary vessels at MR imaging. Segmental enlargement or distortion of the cord was present in three patients (Fig 3). Areas of intramedullary changes in signal intensity that were related to previous hemorrhages and appeared as areas of high signal intensity in T1- and T2-weighted images were seen in two patients. They were associated with extensive hyperintensity in T2-weighted images and contrast enhancement of the cord in the patient with type 1 perimedullary arteriovenous fistula.

Dural Arteriovenous Fistulas after Treatment
MR angiography.—Early posttreatment MR angiography showed complete or partial persistence of flow in the perimedullary vessels in seven (37%) of the 19 patients treated by means of embolization (Fig 1). Results of DSA and subsequent surgery were used to confirm persistent opening of the fistula in all seven of them. Early and follow-up MR angiography showed disappearance of flow in the perimedullary vessels in 22 (73%) of the 30 patients with dural arteriovenous fistula and, notably, in all those who had undergone surgery. Posttreatment DSA in one patient was performed to confirm cure after surgery.

In one additional patient treated with the endovascular approach, early MR angiography showed disappearance of flow in perimedullary vessels, which, however, reappeared at follow-up (Fig 2). DSA was performed to confirm reopening of the fistula (Fig 2).

MR imaging.—In the seven patients with early demonstration of failure of endovascular treatment at MR angiography, MR imaging features were unchanged with respect to the pretreatment MR imaging examinations. In particular, cord swelling and changes in signal intensity were present in seven, cord contrast enhancement was present in four, perimedullary vessels were present on T2-weighted images in five, and enhanced vessels were present on postcontrast T1-weighted images in five. In the patient with reopening of the fistula, extension of cord hyperintensity on T2-weighted images and of contrast-enhanced perimedullary vessels were decreased on early posttreatment follow-up images but increased on follow-up images (Fig 2).

At the final posttreatment MR imaging examination in the 22 patients with successful treatment of the dural arteriovenous fistula as indicated by MR angiographic findings and clinical outcome (described later), the spinal cord appeared thinned in two patients, of normal size in 16, and still mildly enlarged in four (18%). Normalization of spinal cord volume was observed as early as 10 days after treatment. Signal hyperintensity in T2-weighted images had completely disappeared in three (14%) of 22 patients, was reduced in extension in 12, and was unchanged in seven. The minimum interval since treatment that led to normalization of cord signal intensity in T2-weighted images was 6 months.

In the 21 patients in this group initially exhibiting spinal cord contrast enhancement, the enhancement had disappeared in five patients (24%), was reduced in extension and conspicuity in 14, and was unchanged in two. Disappearance of cord contrast enhancement occurred as early as 3 weeks after treatment.

Perimedullary vessels appearing as a serpentine area of signal void were no longer appreciable in any patient. In three patients, serpentine structures of intermediate signal intensity were observed. Perimedullary vessel contrast enhancement was still observed in seven (54%) of 13 patients, whereas it had disappeared in six. In three patients, MR imaging showed complications of surgery (one pseudomeningocele and one dural adhesion) or endovascular therapy (one infarction of the posterior spinal artery).

Clinical outcome.—The condition of seven patients with early detection of failure of endovascular treatment remained clinically stable after the procedure. The patient with recurrence experienced prompt improvement of symptoms after therapy, which was followed by deterioration starting 2 months later.

Of the 22 patients with successful treatment of the fistula (as determined at MR angiography and, in one, at DSA), three had had completely recovered, 14 had improved, and five were unchanged.

Intradural Arteriovenous Malformations after Treatment
MR angiography.—In two patients with intramedullary arteriovenous malformations, early posttreatment MR angiography demonstrated mild (n = 1) and marked (n = 1) reduction of flow in the nidus, with flow restoration in subsequent follow-up examinations, which was confirmed at DSA in both (Fig 3). In two patients with perimedullary arteriovenous fistula, early and follow-up MR angiographic results documented disappearance of flow in perimedullary vessels, which was confirmed at DSA in the patient with the type 3 fistula.

MR imaging.—In the patient with thoracic intramedullary arteriovenous malformation treated with embolization with liquid adhesive, the serpentine and punctate areas of signal void within and at the surface of the cord exhibited an intermediate signal intensity at the early follow-up examination. Flow restoration at subsequent follow-up examinations was associated with reappearance of signal void in the same areas. No appreciable differences with the pretreatment findings were observed at the posttreatment MR imaging examinations in the patient with cervical intramedullary arteriovenous malformation treated with embolization with particles.

In the patient with type 1 perimedullary arteriovenous fistula, MR imaging findings were unchanged as compared with pretreatment examination results. In the patient with type 3 perimedullary arteriovenous fistula treated with embolization with liquid adhesive and coil, a round area exhibiting high signal intensity in T1- and T2-weighted images, which is consistent with thrombus, replaced the intramedullary area of signal void of the pretreatment MR examination. Linear areas of signal void within this area and image distortion due to metal artifacts in the left paraspinal region were also present.

Clinical outcome.—Three patients were unchanged, whereas the patient with type 3 perimedullary arteriovenous fistula had improved.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
REFERENCES
 
Because of the different clinical presentation, pathophysiology, and treatment of the two conditions, we shall consider dural arteriovenous fistulas with perimedullary venous drainage separately from intradural arteriovenous malformations. Spinal dural arteriovenous fistulas are the most common spinal vascular lesions in adults and elderly subjects, with a significant male predominance (9–11). Intracranial dural arteriovenous fistulas with perimedullary venous drainage can be considered a related disorder (11–13).

Dural arteriovenous fistula with perimedullary venous drainage typically manifests as a progressive myelopathy (7). It is usually a low-flow lesion fed by meningoradicular arteries and is located in the dura at the craniospinal junction or between T3 or T4 and S2 (9–11).

The main pathophysiologic mechanism responsible for neurologic deficits entails hypertension in the perimedullary venous plexus owing to intradural venous drainage of the fistula via a single dilated intradural vein (14,15). The perimedullary venous hypertension reduces the intramedullary arteriovenous pressure gradient and ultimately produces a congestive myelopathy, which progresses from edema to ischemia, infarction, and necrosis (14,15).

Interruption of the connection of the fistula with the perimedullary venous plexus halts disease progression and can be accomplished either with surgical ligation or excision of the intradural draining vein (11,16,17) or with occlusion of the fistula and the proximal portion of the draining vein by embolization with liquid adhesive (10,18). Occlusion of the fistula with particles is associated with recurrence (19,20).

Pretreatment MR imaging and MR angiographic findings in the series of patients with spinal vascular malformations in our study are in agreement with those of previous reports (1–6,21–26). In particular, they confirm the superiority of MR angiography over MR imaging for depiction of the abnormal perimedullary vessels associated with dural arteriovenous fistulas. This is demonstrated by the appreciation with MR angiography of perimedullary vessels that were missed at T2-weighted and contrast-enhanced T1-weighted MR imaging in nine patients and by the consistently greater extension of the abnormal vessels at MR angiography than at T2-weighted MR imaging and contrast-enhanced T1-weighted MR imaging.

As pointed out by Cognard et al (18), clinical response to endovascular or surgical therapy of dural arteriovenous fistula is not a reliable indicator of effective or failed treatment. Patients may not improve notwithstanding effective treatment, especially if their disease duration is of many years. On the other hand, failure of endovascular therapy has to be ascertained quickly to proceed with surgery and avoid further clinical deterioration.

Since it is often difficult to ascertain if the liquid adhesive has arrived in the intradural draining vein during therapeutic DSA (18), until now, repeated (15 days to 2 months) DSA has been the only means to evaluate the effectiveness of treatment of dural arteriovenous fistulas. Noninvasive techniques are desirable. Immediate postembolization computed tomography was proposed for depiction of the distribution of the liquid adhesive mixed with iodized oil in the spinal compartment (18).

As confirmed in our series, MR imaging has some limitations for this purpose (18,25,27–29). The resolution of spinal cord changes in signal intensity and contrast enhancement at MR imaging in successfully treated dural arteriovenous fistula is slow and often incomplete. Moreover, we noted contrast-enhanced perimedullary vessels in seven patients after effective treatment of the fistula. This finding, which possibly corresponds to enhancement of the dilated perimedullary venous plexus (15,30), suggests caution in using the presence or lack of perimedullary contrast-enhanced vessels as a possible feature in assessing the results of treatment.

We performed MR angiography to evaluate treatment of dural arteriovenous fistula with perimedullary venous drainage, with the premise that MR angiography is a flow-sensitive technique ideal to monitor flow changes in the abnormal vessels after therapy. This feature of the technique was particularly valuable in the depiction of residual flow in perimedullary vessels in seven patients with failed embolization with liquid adhesive and in demonstration of lack of significant flow in contrast-enhanced vessels in the seven patients with effective treatment (described earlier). The only possible deception was in the case of the patient in whom early posttreatment MR angiography failed to help identify a small residual dural fistula that was clearly identified at follow-up. Lack of flow in perimedullary vessels at early posttreatment MR angiography in this patient could reflect insufficient flow sensitivity or spatial resolution of the technique.

Although we suggest caution in substituting MR angiography for posttreatment DSA, it is noteworthy that all seven cases of treatment failure suggested at MR angiography in our study were confirmed at DSA. On the other hand, the long-lasting negative results at posttreatment MR angiography corresponded to clinical improvement and at least partial resolution of myelopathy at MR imaging in 22 patients. In the patient with reopening of a small residual fistula after treatment, DSA could have permitted earlier identification of treatment failure as compared with MR angiography. Overall, our results with posttreatment MR angiography in patients with dural arteriovenous fistula and perimedullary venous drainage help to confirm the results of preliminary reports (3,17) in patients with spinal dural arteriovenous fistula examined with a 3D time-of-flight MR angiographic technique.

Intradural arteriovenous malformations manifest in children or young adults with spinal stroke related to vessel rupture and intramedullary or perimedullary hemorrhage or with relapsing remitting clinical course, which are possibly due to compression by a nidus or by aneurysms, vascular thrombosis, posthemorrhagic arachnoiditis, or "steal phenomenon" (7,15). These lesions are generally low-resistance high-flow lesions fed by the anterior or posterior radiculomedullary arteries (7,8,15).

Except in cases of perimedullary arteriovenous fistulas fed by posterior radiculomedullary arteries, and of intramedullary arteriovenous malformations of the glomus type located in the cervical spinal cord (15,21), surgery is generally avoided in patients with intradural arteriovenous malformations because of the risk of ischemia in the territory of the anterior arterial spinal axis. The endovascular approach may be pursued to decrease flow in the nidus of intramedullary arteriovenous malformations with embolization with particles or liquid adhesive. The former are safer but have a transient effect. Occlusion of intermediate and giant perimedullary fistulas can be accomplished with endovascular therapy by using balloons or coils and liquid adhesive (8).

MR imaging can be performed to monitor reduction of flow in the nidus and perimedullary vessels, which manifests as disappearance of the areas of flow-related signal void (21,23). Although Hasuo et al (29) reported the usefulness of postcontrast MR imaging in the evaluation of treatment of intradural arteriovenous malformations, we believe that MR angiography, which we always performed after administration of contrast material, is more suitable for such a purpose. Our results, although limited to four patients, support a role for MR angiography in the assessment of treated intradural arteriovenous malformations and reduction of the number of follow-up DSA examinations and the amount of radiation exposure in these young adults.

In this study, we used different MR angiographic techniques, and it is reasonable to foresee that the results obtained by ourselves and others may be influenced by the flow sensitivity of the technique used. Although contrast-enhanced 3D time-of-flight MR angiography is especially suited to demonstrate low-flow vessels, including normal medullary veins and veins draining dural arteriovenous fistulas (3,31), fast contrast-enhanced MR angiography is especially suited to demonstration of the high-flow (ie, arterial) components of the malformations (6). In our current MR angiographic protocol, we combine a fast contrast-enhanced time-resolved technique with phase-contrast acquisitions at low or high velocity of encoding, selected according to the time of depiction of the abnormal intrathecal vessels. This approach resolves difficulties in a priori selection of the velocity of encoding for phase-contrast MR angiography and reasonably allows detection of all of the types of spinal vascular malformations.

In conclusion, our results indicate that MR angiography has to be added to MR imaging in the assessment of treatment of spinal vascular lesions, insofar as it provides more robust information about residual or recurrent flow in the abnormal peri- and intramedullary vessels. An early follow-up examination is recommended, especially in patients with dural arteriovenous fistula treated with the endovascular approach. This early MR angiographic examination enables prompt detection of treatment failure and serves as a reference for follow-up examinations.

MR imaging remains fundamental for identification of complications of endovascular and surgical therapy, and it may be performed to monitor the slow and often incomplete recovery from congestive myelopathy in patients with successfully treated dural arteriovenous fistula. MR angiography is a promising modality to substitute for DSA in the posttreatment evaluation of spinal vascular malformations.


   FOOTNOTES
 
Abbreviations: DSA = digital subtraction angiography, 3D = three-dimensional, 2D = two-dimensional

Author contributions: Guarantor of integrity of entire study, N.V.; study concepts and design, M.M.; literature research, L.S.P.; clinical studies, G.F., S.M.; data acquisition, M. Cosottini, M. Cellerini, G.F., N.Q.; data analysis/interpretation, M.M., M. Cosottini; manuscript preparation, L.G.; manuscript definition of intellectual content and editing, M.M.; manuscript revision/review, C.B.; manuscript final version approval, N.V.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 

  1. Gelbert F, Guichard JP, Mourier KL, et al. Phase-contrast MR angiography of vascular malformations of the spinal cord at 0.5 T. J Magn Reson Imaging 1992; 2:631-636.[Medline]
  2. Mascalchi M, Quilici N, Ferrito G, et al. MR angiography of spinal vascular malformations. AJNR Am J Neuroradiol 1995; 16:289-297.[Abstract]
  3. Bowen BC, Fraser K, Kochan JP, Pattany PM, Green BA, Quencer RM. Spinal dural arteriovenous fistulas: evaluation with MR angiography. AJNR Am J Neuroradiol 1995; 16:2029-2043.[Abstract]
  4. Mascalchi M, Quilici N, Ferrito G, et al. Identification of the feeding arteries of spinal vascular lesions via phase-contrast MR angiography with three-dimensional acquisition and phase display. AJNR Am J Neuroradiol 1997; 18:351-358.[Abstract]
  5. Mascalchi M, Cosottini M, Ferrito G, Quilici N, Bartolozzi C, Villari N. Contrast-enhanced time-resolved MR angiography of spinal vascular malformations. J Comput Assist Tomogr 1999; 23:341-345.[Medline]
  6. Binkert CA, Kollias SS, Valavanis A. Spinal cord vascular disease: characterization with fast three-dimensional contrast-enhanced MR angiography. AJNR Am J Neuroradiol 1999; 20:1785-1793.[Abstract/Free Full Text]
  7. Rosenblum B, Oldfield EH, Doppman JL, Di Chiro G. Spinal arteriovenous malformations: a comparison of dural arteriovenous fistulas and intradural AVMs in 81 patients. J Neurosurg 1987; 67:795-802.[Medline]
  8. Gueguen B, Merland JJ, Riche MC, Rey A. Vascular malformations of the spinal cord: intrathecal perimedullary arteriovenous fistulas fed by medullary arteries. Neurology 1987; 37:969-979.[Abstract/Free Full Text]
  9. Kendall BE, Logue V. Spinal epidural angiomatous malformations draining into intrathecal veins. Neuroradiology 1977; 13:181-189.[Medline]
  10. Merland JJ, Riche MC, Chiras J. Les fistules artério-veineuses intra-canalaires, extra-médullaires à drainage veineux médullaire. J Neuroradiol 1980; 7:271-320.[Medline]
  11. Symon L, Kuyama H, Kendall B. Dural arteriovenous malformations of the spine: clinical features and surgical results in 55 cases. J Neurosurg 1984; 60:238-247.[Medline]
  12. Wrobel CJ, Oldfield EH, Di Chiro G, Tarlov EC, Baker RA, Doppman JL. Myelopathy due to intracranial dural arteriovenous fistulas draining intrathecally into spinal medullary veins. J Neurosurg 1988; 69:934-939.[Medline]
  13. Versari PP, D’Alimberti G, Talamonti G, Branca V, Boccardi E, Collice M. Progressive myelopathy caused by intracranial dural arteriovenous fistula: report of two cases and review of the literature. Neurosurgery 1993; 33:914-919.[Medline]
  14. Aminoff MJ, Barnard RO, Logue V. The pathophysiology of spinal vascular malformations. J Neurol Sci 1974; 23:255-263.[Medline]
  15. Crisciulo GR, Rothbart D. Vascular malformations of the spinal cord: pathophysiology, diagnosis, and management. Neurosurg Q 1992; 2:77-98.
  16. Afshar JKB, Doppman JL, Oldfield EH. Surgical interruption of intradural draining vein as curative treatment of spinal dural arteriovenous fistulas. J Neurosurg 1995; 82:196-200.[Medline]
  17. Lee TT, Gromelski EB, Bowen BC, Green BA. Diagnostic and surgical management of spinal dural arteriovenous fistulas. Neurosurgery 1998; 43:242-247.[Medline]
  18. Cognard C, Miaux Y, Pierot L, Weill A, Martin N, Chiras J. The role of CT in evaluation of the effectiveness of embolization of spinal dural arteriovenous fistulae with N-butyl cyanoacrylate. Neuroradiology 1996; 38:603-608.[Medline]
  19. Hall WA, Oldfield EH, Doppman JL. Recanalization of spinal arteriovenous malformations following embolization. J Neurosurg 1989; 70:714-720.[Medline]
  20. Nichols DA, Rufenacht DA, Jack CR, Jr, Forbes GS. Embolization of spinal dural arteriovenous fistula with polyvinyl alcohol particles: experience in 14 patients. AJNR Am J Neuroradiol 1992; 13:933-940.[Abstract]
  21. Doppman JL, Di Chiro G, Dwyer AJ, Frank JL, Oldfield EH. Magnetic resonance imaging of spinal arteriovenous malformations. J Neurosurg 1987; 66:830-834.[Medline]
  22. Masaryk TJ, Ross JS, Modic MT, Ruff RL, Selman WR, Ratcheson RA. Radiculomeningeal vascular malformations of the spine: MR imaging. Radiology 1987; 164:845-849.[Abstract/Free Full Text]
  23. Minami S, Sagoh T, Mishimura K, et al. Spinal arteriovenous malformation: MR imaging. Radiology 1988; 169:109-115.[Abstract/Free Full Text]
  24. Terwey B, Becker H, Thron AK, Vahldiek G. Gadolinium-DTPA enhanced MR imaging of spinal dural arteriovenous fistulas. J Comput Assist Tomogr 1989; 13:30-37.[Medline]
  25. Larrson EM, Desai P, Hardin CW, Story J, Jinkins JR. Venous infarction of the spinal cord resulting from dural arteriovenous fistula: MR imaging findings. AJNR Am J Neuroradiol 1991; 12:739-743.[Abstract]
  26. Gilbertson JR, Miller GM, Goldman MS, Marsh WR. Spinal dural arteriovenous fistulas: MR and myelographic findings. AJNR Am J Neuroradiol 1995; 16:2049-2057.[Abstract]
  27. Isu T, Iwasaki Y, Akino M, Koyanagi I, Abe H. Magnetic resonance imaging in cases of spinal dural arteriovenous malformation. Neurosurgery 1989; 6:919-923.
  28. Willinsky RA, terBrugge K, Montanera W, Mikulis D, Wallace MC. Posttreatment MR findings in spinal dural arteriovenous malformations. AJNR Am J Neuroradiol 1995; 16:2063-2071.[Abstract]
  29. Hasuo K, Mizushima A, Mihara F, et al. Contrast-enhanced MRI in spinal arteriovenous malformations and fistulae before and after embolization therapy. Neuroradiology 1996; 38:609-614.[Medline]
  30. Hassler W, Thron A, Grote EH. Hemodynamics of spinal dural arteriovenous fistulas: an intraoperative study. J Neurosurg 1989; 70:360-370.[Medline]
  31. Bowen BC, DePrima S, Pattany PM, Marcillo A, Masden P, Quencer RM. MR angiography of normal intradural vessels of the thoracolumbar spine. AJNR Am J Neuroradiol 1996; 17:483-494.[Abstract]



This article has been cited by other articles:


Home page
 Am. J. Neuroradiol.Home page
W.H. Backes and R.J. Nijenhuis
Advances in Spinal Cord MR Angiography
AJNR Am. J. Neuroradiol., April 1, 2008; 29(4): 619 - 631.
[Abstract] [Full Text] [PDF]

Home page
 BrainHome page
K. Jellema, C. C. Tijssen, and J. v. Gijn
Spinal dural arteriovenous fistulas: a congestive myelopathy that initially mimics a peripheral nerve disorder
Brain, December 1, 2006; 129(12): 3150 - 3164.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Neuroradiol.Home page
R.J. Nijenhuis, M. Mull, J.T. Wilmink, A.K. Thron, and W.H. Backes
MR Angiography of the Great Anterior Radiculomedullary Artery (Adamkiewicz Artery) Validated by Digital Subtraction Angiography
AJNR Am. J. Neuroradiol., August 1, 2006; 27(7): 1565 - 1572.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Neuroradiol.Home page
R.J. Nijenhuis, M.J. Jacobs, J.M.A. van Engelshoven, and W.H. Backes
MR Angiography of the Adamkiewicz Artery and Anterior Radiculomedullary Vein: Postmortem Validation
AJNR Am. J. Neuroradiol., August 1, 2006; 27(7): 1573 - 1575.
[Abstract] [Full Text] [PDF]

Home page
 StrokeHome page
J. M. C. van Dijk, K. G. TerBrugge, R. A. Willinsky, R. I. Farb, and M. C. Wallace
Multidisciplinary Management of Spinal Dural Arteriovenous Fistulas: Clinical Presentation and Long-Term Follow-Up in 49 Patients
Stroke, June 1, 2002; 33(6): 1578 - 1583.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Neuroradiol.Home page
E. Saraf-Lavi, B. C. Bowen, R. M. Quencer, E. M.L. Sklar, A. Holz, S. Falcone, R. E. Latchaw, R. Duncan, and A. Wakhloo
Detection of Spinal Dural Arteriovenous Fistulae with MR Imaging and Contrast-Enhanced MR Angiography: Sensitivity, Specificity, and Prediction of Vertebral Level
AJNR Am. J. Neuroradiol., May 1, 2002; 23(5): 858 - 867.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
R. I. Farb, J. K. Kim, R. A. Willinsky, W. J. Montanera, K. terBrugge, J. A. Derbyshire, J. M. C. van Dijk, and G. A. Wright
Spinal Dural Arteriovenous Fistula Localization with a Technique of First-Pass Gadolinium-enhanced MR Angiography: Initial Experience
Radiology, March 1, 2002; 222(3): 843 - 850.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mascalchi, M.
Right arrow Articles by Villari, N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mascalchi, M.
Right arrow Articles by Villari, N.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE