Interesting SAH case

[txterp98]

---

Members do not see this ad.

Hey everyone, I'm an ER attending only a couple of years out of residency. I had a pretty interesting case that I was wondering if some of you senior residents/attendings could maybe help share your understanding with me.

A 50-ish yo male went to ER #1 about 2 hours out from a thunderclap onset of a headache. Headache's worst of his life, also associated with sharp lower back pain. ER #1 gets a noncontrasted Head CT which is read to be normal. Time from presumed aneursymal rupture to time of CT is around the 3 1/2 hr to 4 hr timeframe. No LP is performed. Patient is told he has a migraine and a back sprain.

Patient comes to my ER 2 1/2 days later where I see him. Patient's c/o ongoing HA, ongoing lower back pain, difficulty voiding. He's completely neurologically intact. I didn't check a post-void residual urine, though on hindsight, I possibly should've. At the time, though, no saddle anesthesia or urinary incontinence. His noncontrast Head CT is read by our neuroradiologist as normal. My LP is classic non-clearing Kool-Aid appearing CSF in a very easy, definitely atraumatic tap. I talk with the neurosurgeon to get him on board and then the interventional neuroradiologist takes him for a 4-vessel cerebral angiogram.

Angiogram is without hemorrhage or aneurysm. Neurology is consulted, calls it a traumatic LP (man, that got me going!) and orders and MRI Brain w/ Gad, MRI L-spine with/without Gad. No MRA is ordered. MRI Brain reveals occipital subarachnoid blood; MRI L-spine reveals diffuse subarachnoid hemorrhage including partial layering around L5/S1 rootlets. So basically, the MRI's confirmed 2 things I had clinically suspected - ruptured aneurysm and possibly early signs of cauda equina symptoms secondary to coagulating blood in the dependent portion of his spine. Fortunately at this time, he still hasn't developed overflow incontinence and has no motor/sensory abnormalities.

So the plan now is in 1 week for a cerebral/spinal angiogram in 1 week to assess for aneurysms/AVM's anywhere along his brain/neural axis.

I had a few questions, if you don't mind me asking. I promise you, I have tried to do a literature review on this as well.

1) Any idea what the false negative rate is on a first attempt angiogram for suspected aneurysm? From my review, I guess clotted aneurysms, posterior aneurysms, aneurysms < 5mm, and vasospasm all decrease the sensitivity?

2) In general, what is the conventional wisdom on sensitivities for detecting a SAH with a noncontrasted Head CT, CTA Brain, MRA brain, and conventional angiogram. On my review, I'm getting 60-95% for noncontrast Head CT based on time of presentation; CTA around 50% for posterior aneurysms vs 85% anterior aneurysms; MRA around 90%; and I haven't found any numbers for conventional angiogram.

3) I'm sorry if this comes across naively, but is conventional angiography the same thing as a digital subtraction angiography?

Obviously, the answers here don't change the ER practice - noncontrast Head CT, LP if negative, discuss with NSG/Neurorads if suspicious. I'm just hoping to add to my knowledge base on this topic.

Thanks for your time.

 

[f_w]

My money is on dural AVM somewhere along the upper spine. Supply for that doesn't have to come from the 6 vessels checked during a 4-vessel angio (alternative is something like a PICA aneurysm masked by early vasospasm due to the delayed presentation).

No LP is performed. Patient is told he has a migraine and a back sprain.

Famous last words.

Neurology is consulted, calls it a traumatic LP (man, that got me going!) and orders and MRI Brain w/ Gad, MRI L-spine with/without Gad.

Typical neurologic reflex behaviour (the same neurologist will probably pawn off his own LPs to IR or anesthesia so he doesn't have to interrupt his clinic schedule).

So the plan now is in 1 week for a cerebral/spinal angiogram in 1 week to assess for aneurysms/AVM's anywhere along his brain/neural axis.

Those are fun. 20+ pairs of spinal arteries. Pack a lunch and hang a bucket of contrast.

1) Any idea what the false negative rate is on a first attempt angiogram for suspected aneurysm? From my review, I guess clotted aneurysms, posterior aneurysms, aneurysms < 5mm, and vasospasm all decrease the sensitivity?

Without being able to give you a reference, it is somewhere between 5-10%. Your patient is the typical setup for a miss. 3 days out and posterior fossa. To reduce the risk of traumatizing the the verts, some neuroradiologists only inject the dominant vert and rely on reflux down the non-dominant vert to fill the PICA. This usually works, but sometimes the bear will get you.

2) In general, what is the conventional wisdom on sensitivities for detecting a SAH with a noncontrasted Head CT

I assume a 5% false negative.

, CTA Brain, MRA brain, and conventional angiogram. On my review, I'm getting 60-95% for noncontrast Head CT based on time of presentation; CTA around 50% for posterior aneurysms vs 85% anterior aneurysms; MRA around 90%;

Sounds about right.

3) I'm sorry if this comes across naively, but is conventional angiography the same thing as a digital subtraction angiography?

'conventional angiography' was an arduous process where you had a fluoro unit to place your catheters/wires and then recorded the actual contrast run on a 'rapid film changer' (e.g. a Siemens 'Puck') using regular x-ray film. After a contrast run, 2-3 techs would go to the darkroom and develop 20 sheets of film (40 sheets in a biplane run). A 6 vessel cerebral angio was a 2 hour affair because you had to wait for images of every run to be developed to decide whether you had to repeat it.

Downside: you needed oodles of contrast and time.

Upsides: 1ms temporal resolution (no motion artifacts), exquisite spatial resolution (regular x-ray film). Nothing beats a 'conventional' angio to rule out cerebral vasculitis.

I can proudly say that I have seen a 'conventional angiogram' done once in my life. But that sort of dates me (I am not old enough for pneumoencephalograms though).

DSA is essentially a fancy fluoro unit that is able to record frames and digitally process them by using subtraction and digital filters. The disadvantage is lower spatial resolution compared with the old-fashioned conventional angiogram. But the massively reduced hassle factor has led to it replacing conventional angio.

Obviously, the answers here don't change the ER practice - noncontrast Head CT, LP if negative, discuss with NSG/Neurorads if suspicious.

If I had a buck for every time this concept is violated, I wouldn't have to go to work.

 

[txterp98]

Thanks for the help f_w. I'll be sure to post back what they find in a few days. Good call on the possibility of a dural AVM in the upper spine. While everyone's in the waiting stage, the MRI of the rest of the spine has shown what appears to be a dural hemorrhage at the upper thoracic levels.

Thanks, again.

 

[Docxter]

One point is that nowadays CTA is much more sensitive than MRA. The studies that you may have come across are older ones with poor, older CTA techniques.

Another thing is that you mentioned, "So basically, the MRI's confirmed 2 things I had clinically suspected - ruptured aneurysm ...." Actually the studies you mention do not confirm your suspicion of an aneurysm, but that some sort of vascular lesion (not necessarily and probably not an aneurysm) has bled.

fw is right, but in this day and age, conventional cerebral angiography is pretty much the same as digital subtraction angiography. "Conventional" is the term used in distinction to CTA and MRA, and denotes the use of invasive catheter angiography.

 

[txterp98]

Actually the studies you mention do not confirm your suspicion of an aneurysm, but that some sort of vascular lesion (not necessarily and probably not an aneurysm) has bled.

Good point. Everyone who's involved in this patient's care is pretty curious about tomorrow's cerebrospinal angiogram to hopefully find out the type and whereabouts of the lesion.

 

[f_w]

Good point. Everyone who's involved in this patient's care is pretty curious about tomorrow's cerebrospinal angiogram to hopefully find out the type and whereabouts of the lesion.

Might want to hang a liter of bicarb in the morning (150mEq/l), your patient is going to see a lot of contrast (and don't mind the slight blue glow of cherenkoff radiation when he returns to the floor 😉).

 

[hans19]

Good point. Everyone who's involved in this patient's care is pretty curious about tomorrow's cerebrospinal angiogram to hopefully find out the type and whereabouts of the lesion.

Suspense is killing me... so where was it?