Cord Infarct
- Spinal cord infarct is a stroke of the cord — usually the anterior spinal artery territory, which feeds the front two-thirds of the cord.
- It comes on fast (minutes to hours), which is the single biggest clue separating it from inflammatory mimics that build over days.
- The MRI signature is a sausage of T2-bright signal in the central/anterior cord, classically the "owl eyes" of the anterior horns on axial images, with restricted diffusion early.
- Look for a reason: aortic surgery, aortic dissection, severe hypotension, or a vertebral body infarct sitting right next to the cord lesion.
- The big trap is calling it transverse myelitis and reaching for steroids when the cord has actually been starved of blood.
The cord runs blood like everything else in the body, and like everything else, it can have a stroke. We just don't talk about it as much because the brain hogs all the attention. But when the cord's blood supply fails, the result is sudden, dramatic, and — on the wrong day — permanent.
The whole story hinges on one piece of plumbing: the cord is fed mostly from the front, by a single skinny vessel running down its length. When that vessel quits, the front of the cord goes dark.
The plumbing (why the front fails first)
Picture the cord in cross-section as a butterfly of gray matter floating in white matter. Now imagine its water supply: one main pipe down the front midline — the anterior spinal artery — feeding roughly the anterior two-thirds, and two smaller pipes down the back covering the rest.
Here's the cruel design flaw. That single anterior pipe is fed along the way by little tributaries called radiculomedullary arteries, and there are stretches with embarrassingly few of them. The mid-thoracic cord is the classic dry zone — a long garden hose supplied by one faucet at the top and one near the bottom, with a parched stretch in the middle. Drop the pressure anywhere upstream and that watershed middle is the first to wilt.
So when people say anterior spinal artery syndrome, they mean exactly this: the front pipe failed, the front of the cord died, and the back — sensing position and vibration through those posterior pipes — often gets spared.
Clinically that split is the giveaway: sudden loss of motor function and pain/temperature sensation below the level, while the patient can still feel where their toes are and whether something is vibrating. Front pipe down, back pipes fine.
What it looks like on MRI
MRI is the test, and the cord lesion has a recognizable shape because it follows the dead territory, not a random blob.
On sagittal T2, you get a vertically oriented "pencil-like" streak of bright signal in the central/anterior cord, spanning a few segments. On axial T2, the most quoted pattern is bright signal in the paired anterior gray matter — the "owl eyes" or "snake eyes," two glowing dots where the anterior horn cells live, because gray matter is the hungriest tissue and dies first.
Then there's diffusion. Just like a brain stroke, an acute cord infarct restricts diffusion — bright on DWI, dark on ADC — because the starving cells swell and water stops moving freely. If you want the physics of why dead tissue traps water, that lives over in diffusion (DWI/ADC) physics. Cord DWI is technically fiddly and not always available, so don't panic if it's not on your study — but when it's clean and positive, it's gold.
The detail everyone forgets to look for
Look at the vertebral bodies. The same little radiculomedullary arteries that feed the cord often share a source with the arteries feeding the adjacent vertebral marrow. So a cord infarct may sit right next to a vertebral body infarct — a wedge of marrow that lights up on T2/STIR and restricts diffusion too.
A bright, infarcted vertebral body next to the cord lesion is a strong vote for ischemia over inflammation. When the bone and the cord went down together, blame the blood supply.
Why it happened (always ask)
A cord infarct is rarely random. Hunt for the cause, because the cause changes the whole conversation:
| Setting | Why the cord starved |
|---|---|
| Aortic surgery / endovascular aortic repair | The radiculomedullary feeders branch off the aorta; clamping or covering them cuts supply. |
| Aortic dissection | A dissection flap can shear off or occlude those feeders. |
| Severe hypotension / cardiac arrest | Global low pressure drops out the watershed cord first. |
| Embolic / fibrocartilaginous embolism | Disc material embolizing into cord arteries — rare, but real. |
The mimic you must not miss
The trap is transverse myelitis. Both give a long T2-bright cord and an acute-ish neuro deficit, and both can land on your worklist at 2 a.m.
The tempo is your best friend. Infarct is abrupt — deficit maxes out in minutes to hours. Transverse myelitis builds over days, more often shows central cord T2 with patchy enhancement, and doesn't restrict diffusion the way an acute infarct does. Call infarct "myelitis," give steroids, and you've treated a stroke with anti-inflammatories — wrong door entirely.
And before you commit to either, make sure something hard isn't squeezing the cord from outside. A quick scan for mass effect rules out cord compression, which is a different — and often surgical — emergency.
The takeaway
If a patient drops their legs in minutes, you find a pencil of anterior T2 brightness with owl-eye gray matter and early restricted diffusion, and there's an aortic event or a crash in the history — think stroke of the cord, not inflammation. Fast onset, anterior territory, restricted diffusion, and a reason. That's the whole package.