Neurointervention · IR Playbook

Spinal Angiography (SpDSA)

Spinal angiography is selective catheter-based digital subtraction angiography of the arteries supplying the spinal cord, dura and spine. It is the reference standard for detecting and characterizing spinal arteriovenous shunts and for mapping feeders, normal cord supply and venous drainage before treatment. Modern expert series show a low complication rate, but spinal cord or cerebral ischemia is a rare potentially devastating risk; diagnostic quality depends on a complete, technically adequate and correctly labeled examination.

Learning objectives

• Select appropriate indications and define whether a targeted or complete spinal angiogram is required.
• Recognize the segmental, radiculomedullary, radiculopial and longitudinal arterial anatomy of the spinal cord.
• Perform safe bilateral selective DSA with accurate vertebral-level labeling and adequate venous-phase imaging.
• Describe a spinal vascular lesion by its exact level, arterial supply, shunt or nidus and venous drainage.
• Recognize technical causes of a false-negative study and the major procedure-related complications.

Indications

• MRI or MRA findings suspicious for a spinal dural arteriovenous fistula, epidural arteriovenous fistula, intradural arteriovenous malformation or perimedullary arteriovenous fistula.
• Unexplained spinal subarachnoid hemorrhage, hematomyelia or selected spinal cord infarction when an underlying vascular lesion remains suspected.
• Definitive angioarchitectural mapping before endovascular or surgical treatment of a spinal vascular lesion.
• Assessment for residual, recurrent or additional shunting after embolization or surgery.
• Preoperative mapping of a hypervascular spinal or paraspinal tumor and identification of normal spinal cord arterial supply at risk.
• Selected preoperative localization of critical radiculomedullary supply before complex spinal or thoracoabdominal aortic surgery.

Relative contraindications

• Clinically significant coagulopathy or thrombocytopenia.
• Severe iodinated-contrast hypersensitivity.
• Renal insufficiency, especially when a complete study may require a large contrast volume.
• Uncontrolled severe hypertension, hypotension or decompensated heart failure.
• Severe aortic atheroma, occlusive disease, tortuosity or other anatomy that substantially increases catheterization risk.
• Inability to cooperate or remain motionless without appropriate anesthesia support.

Equipment

• Ultrasound machine with sterile cover and 21 G micropuncture access set with 0.018 inch wire.
• 4–5 Fr arterial sheath, typically short for transfemoral access.
• 0.035 inch J-tip wire and angled hydrophilic guidewire.
• 4 Fr selective catheters such as Cobra 2, Headhunter, Mikaelsson, RDC and Simmons 1 or 2.
• Optional 1.5–2.8 Fr neurovascular microcatheter and compatible microwire for essential superselective diagnostic imaging.
• Continuously flushed, air-free pressure tubing, manifold, three-way stopcocks and contrast syringes.
• Nonionic low- or iso-osmolar iodinated contrast with a contrast-dose record.
• Biplane DSA system with roadmapping, magnification, collimation and optional rotational angiography or cone-beam CT.
• Standard physiologic monitoring, neurologic emergency drugs and contrast-reaction resuscitation equipment.
• Manual-compression supplies or a compatible arterial closure device.

Procedure steps

1. Review MRI, MRA or CTA at the console, define the suspected level and decide whether the examination must be targeted or complete. Establish a vertebral count from fixed landmarks and document the baseline neurologic examination.Pearl: A noninvasive localizing study reduces unnecessary segmental injections but must not truncate a study when clinical suspicion remains high.
2. Place the patient supine, minimize spinal rotation and secure comfortable immobility. Apply continuous cardiovascular and oxygenation monitoring and use sedation or general anesthesia appropriate to the expected duration and ability to cooperate.Pearl: Motion control is fundamental because misregistration can hide tiny shunts.
3. Use ultrasound-guided common femoral arterial access and insert a 4–5 Fr sheath; use radial access only when anatomy and target coverage are suitable. Connect an air-free continuously flushed system and administer intraprocedural anticoagulation only according to local neuroangiography practice and patient risk.Pearl: Small diagnostic systems reduce arteriotomy size while permitting stable selective catheterization.
4. Count vertebral levels fluoroscopically from an agreed fixed landmark and annotate each injection by side and level. Recheck the count at junctional regions and when ribs, transitional vertebrae or prior hardware create ambiguity.Pearl: Include a recognizable osseous landmark on saved images whenever possible.
5. When the cervical or upper thoracic cord is relevant, selectively examine bilateral vertebral arteries and appropriate ascending cervical and deep cervical branches from the thyrocervical and costocervical trunks. Record normal radiculomedullary or radiculopial supply before interpreting or treating a lesion.
6. Engage right and left intercostal or lumbar arteries sequentially with the catheter tip directed toward the ostium. Use gentle contrast puffs and roadmapping when necessary; maintain a systematic level-by-level sequence to avoid omissions.Pearl: The spinal axis is midline, so an apparently unilateral MRI abnormality does not justify omitting the opposite pedicles.
7. Confirm free antegrade flow and a stable nonwedged catheter position with a small low-pressure test injection. For a typical adult segmental artery, approximately 2 mL/s for 4 mL is a common starting acquisition, adjusted downward for vessel caliber and reflux.Pearl: A diagnostic run should opacify the segmental artery and its branches without forceful reflux.
8. Center and collimate the field, use magnification, suspend respiration when possible and acquire from the arterial phase through capillary and venous drainage. Use additional oblique or lateral projections when vessels overlap or the shunt point is unclear.Pearl: Slow-flow fistulas may only become convincing when early venous filling is compared with an adequately recorded normal venous phase.
9. When evaluating the conus, cauda equina or a study that remains negative, examine appropriate internal iliac branches including iliolumbar and lateral sacral arteries and the median sacral artery. Include the pelvic contributors in the documented completeness map.
10. Use targeted high-magnification DSA, oblique views, rotational angiography or cone-beam CT when necessary to separate the feeder, shunt point, nidus and venous outlet. Reserve superselective microcatheterization for an essential diagnostic or therapeutic question in expert hands.Pearl: Three-dimensional acquisitions are adjuncts to, not substitutes for, complete selective DSA.
11. For each lesion, record exact vertebral level and side, arterial feeder, anterior or posterior spinal artery relationship, shunt point or nidus, associated aneurysm and venous drainage pattern. If no lesion is found, audit whether every required territory was injected, technically adequate and correctly interpreted.Pearl: Operator-dependent causes dominate false-negative spinal angiograms: a documented lesion may be overlooked, the region may not be examined or an injection may be inadequate.
12. Remove the sheath and achieve hemostasis by manual compression or a compatible closure device. Immediately repeat and document the focused neurologic examination, access-site status and distal pulses.

Complications

• Spinal cord or cerebral ischemia
• Catheter-related arterial dissection or vasospasm
• Access-site hematoma, pseudoaneurysm or retroperitoneal hemorrhage
• Iodinated-contrast hypersensitivity
• Contrast-associated acute kidney injury
• Radiation-induced skin injury

Exam pearls

• Spinal DSA is the reference standard for spinal arteriovenous shunts; MRA and CTA primarily localize and guide it.
• The spinal axis is midline: a reliable examination usually requires bilateral selective injections.
• Never force contrast through a wedged segmental catheter; restore free flow before any diagnostic run.
• A negative study is credible only when the required territories were completely examined, injections were adequate and images were expertly reviewed.
• Always label the exact vertebral level and side and resolve transitional anatomy before reporting or treatment.
• Record the venous phase; early venous filling and delayed perimedullary drainage are central to diagnosing and characterizing arteriovenous shunts.
• Before embolization or surgery, identify whether a normal anterior or posterior spinal artery arises from the same pedicle as the lesion.
• Lower spinal or conus lesions may be supplied by iliolumbar, lateral sacral or median sacral arteries; a study ending at the lumbar arteries is not always complete.

Viva questions

What is the gold-standard investigation for a suspected spinal vascular malformation?

Selective catheter spinal DSA is the reference standard. MRI and time-resolved MRA or CTA can suggest and localize a lesion, but DSA best defines the feeder, shunt point or nidus, normal cord supply and venous drainage.

What constitutes a complete spinal angiogram?

It is a clinically appropriate bilateral selective examination of all potential arterial contributors, including vertebral and cervical branches, intercostal, lumbar, internal iliac or lateral sacral and median sacral arteries as required. Every injection must be technically adequate, correctly level-labeled and recorded through the relevant venous phase.

Why must both sides be injected at each relevant level?

The spinal cord and its venous system are midline structures, and the feeding artery may arise contralateral to the dominant MRI abnormality. Unilateral sampling can therefore miss a dural or perimedullary shunt.

What is the artery of Adamkiewicz?

It is the great anterior radiculomedullary artery, usually left-sided and arising in the lower thoracic or upper lumbar region, which provides a major contribution to the anterior spinal artery. Its origin is variable and must be identified before treating a segmental pedicle.

How do you describe a spinal dural arteriovenous fistula angiographically?

State the exact vertebral level and side, the radiculomeningeal arterial feeder, the shunt at the dural nerve-root sleeve, early drainage into a radicular vein and the direction and extent of congested perimedullary venous drainage. Also document any normal radiculomedullary artery arising from the same pedicle.

Why can spinal angiography be falsely negative?

Correctable causes include failure to inject the lesion-bearing territory, technically inadequate selective injection, wrong level labeling, incomplete venous-phase acquisition and failure to recognize a lesion already visible on the images. High clinical suspicion should trigger a completeness and interpretation audit.

What is the most feared complication and how is it prevented?

Spinal cord infarction with permanent neurologic deficit is the most feared complication. Prevention includes gentle air-free catheter technique, avoidance of wedged or forceful injections, recognition of normal cord supply, maintenance of perfusion and minimizing catheter trauma and thromboembolism.

When is radial access useful?

Radial access is a feasible alternative for selected targeted or thoracolumbar studies and when femoral or aortoiliac anatomy is unfavorable. The operator must ensure that catheter length, support and access geometry allow every required vessel to be examined.

References