14.3 Noninvasive Angiography and Catheter Angiography
Key Takeaways
- Catheter digital subtraction angiography is the gold-standard vessel study for aneurysm, AVM/dural AV fistula, and angiographic vasospasm because spatial and temporal resolution exceed CTA and MRA.
- Diagnostic DSA risks ischemic stroke, catheter-related dissection, and groin hematoma or pseudoaneurysm, in addition to iodinated-contrast kidney injury.
- CTA-negative SAH with a diffuse cisternal (aneurysmal-pattern) blood distribution still needs catheter angiography; classic perimesencephalic SAH has a much lower aneurysm yield.
- Delayed cerebral ischemia after aneurysmal SAH is monitored with CTA or DSA (and perfusion when available) in the day-4-to-14 vasospasm window; angiographic narrowing is not identical to clinical DCI.
- CTA and DSA both deliver iodinated contrast; time-of-flight MRA avoids iodine and gadolinium when the question is large-vessel patency and the kidneys or allergy history are hostile.
Noninvasive Angiography and Catheter Angiography
Quick Answer: CTA is fast and excellent for large-vessel occlusion and most aneurysms. MRA (especially time-of-flight) avoids iodine. Digital subtraction angiography (DSA) is the gold standard for aneurysm, AVM/dural AV fistula, and vasospasm, at the price of stroke, dissection, and groin injury. CTA-negative aneurysmal-pattern SAH still needs DSA. Delayed cerebral ischemia is a day 4–14 CTA/DSA monitoring problem. Iodine hits the kidney; plan it.
Independent OpenExamPrep teaching in this section covers noninvasive angiography and catheter angiography listed under Diagnostic studies and procedural skills in the ABPN Content Specifications (02.A.4–5). This guide is not an ABIM or ABPN product.
Three ways to see the lumen
| Feature | CTA | MRA | DSA |
|---|---|---|---|
| How it works | Timed iodinated bolus, volumetric CT | TOF flow-related enhancement, or gadolinium-enhanced MRA | Intra-arterial iodine, subtracted fluoroscopy |
| Speed / logistics | Minutes; unstable-friendly | Longer; magnet constraints from the MRI section | Angiography suite; arterial puncture |
| Spatial resolution | High; limited at tiny blebs and skull-base bone | Lower than DSA; TOF overestimates stenosis | Highest; dynamic filling and drainage |
| Best at | LVO, most aneurysms >~3 mm, dissection flap on neck CTA | Avoiding iodine; following known large vessels | Aneurysm confirmation, AVM/dAVF architecture, vasospasm, endovascular treatment |
| Contrast / radiation | Iodine + CT dose | TOF: neither; CE-MRA: gadolinium | Iodine + fluoroscopy dose |
| Kidney | Iodinated load | TOF spares GFR | Iodinated load, often larger if multiple runs |
CTA is the usual code-stroke and thunderclap add-on after NCCT: from the aortic arch through the vertex you see cervical and intracranial arteries, tandem occlusion, and many saccular aneurysms. Dual-energy or bone-subtraction helps at the skull base. CTA under-calls some aneurysms under about 3 mm and can miss slow-filling dural AV fistulas and fine AVM nidus anatomy.
Time-of-flight (TOF) MRA needs no contrast: unsaturated inflowing blood is bright. Slow flow, in-plane flow, and turbulence look like stenosis or occlusion that DSA later disproves. Contrast-enhanced MRA is better for neck vessels and aortic arch origins. Phase-contrast MRA can encode direction and velocity. Pick TOF when iodine and gadolinium are both unattractive and the question is large-vessel patency, not a 2 mm posterior communicating aneurysm.
DSA remains the reference standard when the clinical question is is there a small aneurysm, an AVM, a dAVF, or angiographic vasospasm, and can I treat it in the same sitting? Collaterals, early venous filling, and nidal architecture are DSA language. CTA and MRA are screening and follow-up tools; they do not retire the catheter for those gold-standard indications.
CTA versus MRA versus DSA in common stems
Acute hemiparesis, NCCT without blood: CTA (often with CTP from the prior section) looks for LVO. MRA is an alternative if iodine is forbidden and the patient can do MRI. DSA is not the first diagnostic test in a community ED; it is the treatment room after LVO is found.
Thunderclap, NCCT with cisternal SAH: CTA next looking for aneurysm. If CTA shows a clear saccular aneurysm, the team plans clipping or coiling. If CTA is negative and the blood pattern is diffuse cisternal / sylvian / interhemispheric (aneurysmal pattern), DSA is still indicated. Occult aneurysms, thrombosis in a sac, and operator-dependent CTA misses are the reason. A classic perimesencephalic pattern (blood centered anterior to the midbrain, no peripheral extension) plus a high-quality negative CTA has a low aneurysm yield; many centers still perform at least one DSA, and practice variation is real—do not tell the exam that CTA-negative perimesencephalic blood never gets a catheter, and do not tell it that CTA-negative diffuse SAH is finished.
Suspected AVM or dAVF (pulsatile tinnitus, unexplained ICH in a young patient, dilated vessels): DSA defines feeders, nidus, and venous drainage. CTA/MRA can hint; they do not grade a fistula.
Cervical dissection: CTA or MRA of the neck; DSA if the noninvasive study is inconclusive or endovascular repair is planned. Fat-suppressed T1 MRI of the neck is the intramural-hematoma sequence, not a lumen-only angiogram.
DSA risks: stroke, dissection, groin
Diagnostic cerebral angiography is not a bloodless photograph. Quote the pattern, not a fake exact national rate: ischemic stroke (often embolic from the catheter or from diseased arches) in on the order of a few per thousand to about 1% of diagnostic studies, higher in atherosclerosis, vasculitis, and vasospasm. Permanent disability is less common than transient deficits but is the reason you do not catheterize a low-yield indication. Catheter-induced dissection of the ICA, vertebral artery, or a tortuous arch vessel can cause occlusion or a new flap. Access-site injury after femoral (or radial) puncture includes hematoma, retroperitoneal bleed, pseudoaneurysm, arteriovenous fistula, and limb ischemia. Add iodinated contrast nephropathy, anaphylactoid reaction, and radiation. Death is rare and still on the consent form.
If the stem offers a neurologically intact patient with a tiny, low-risk unruptured aneurysm already well seen on high-quality CTA, the next step is not automatic DSA. If the stem offers CTA-negative diffuse SAH, the stroke risk of DSA is accepted because the risk of a missed ruptured aneurysm is worse.
Delayed cerebral ischemia monitoring with CTA/DSA
After aneurysmal SAH, delayed cerebral ischemia (DCI) clusters around days 4–14, peaking near day 7–10. Angiographic vasospasm is luminal narrowing. Clinical DCI is a new focal deficit or a drop in consciousness not explained by hydrocephalus, rebleeding, seizure, fever, or hyponatremia. The two overlap and are not identical: you can have tight vessels without a deficit and a deficit with only modest narrowing plus a perfusion problem.
Imaging options in that window:
- Transcranial Doppler (a later monitoring chapter) for MCA velocity trends
- CTA as a widely available screen for vessel narrowing
- CTP for a new perfusion deficit that may precede frank infarct on NCCT
- DSA when you need gold-standard spasm grading or intra-arterial therapy
Worked example: day 8 after coiled ACOM aneurysm, new right arm drift, NCCT without hydrocephalus or new blood. CTA shows severe left M1 narrowing. That is vasospasm-related DCI until you treat blood pressure (euvolemia and induced hypertension as in the SAH disease chapter) and consider the suite. A day-2 deficit is too early for typical vasospasm; think hydrocephalus, seizure, or a procedural complication instead.
Do not treat every radiologic spasm with a trip to DSA if the patient is unchanged and perfusion is adequate. Do not ignore a new deficit because “CTA last week was fine”—repeat vascular imaging in the DCI window.
CTA-negative SAH still needs DSA (when the pattern says aneurysm)
Rehearse the blood pattern:
| NCCT SAH pattern | CTA result | Catheter angiography |
|---|---|---|
| Diffuse cisternal, sylvian, interhemispheric, or with ICH/IVH | Negative | DSA indicated; consider repeat DSA if first run is negative and suspicion stays high |
| Isolated convexity sulcal SAH | Negative | Think amyloid, RCVS, thrombosis, endocarditis—not a typical saccular-aneurysm hunt; DSA yield is lower |
| Classic perimesencephalic | High-quality CTA negative | Low aneurysm yield; many protocols still include one DSA |
| CT negative, LP positive (xanthochromia) | CTA negative | Still an occult-aneurysm problem; DSA (and sometimes spine MRI for a spinal source) |
Aneurysms can thrombose after rupture and hide on a single CTA. That is the physiologic excuse for repeat DSA at 1–2 weeks when the first catheter study is negative but the bleed looked aneurysmal. Convexity SAH is a different disease list; blasting every amyloid bleeder with three DSAs is not the lesson.
Contrast and the kidney, again, now with three iodine sources
CTA, CTP, and DSA all use iodinated contrast. A patient can receive a CTP bolus in the ED and a long DSA run the same night. Track cumulative iodine and eGFR.
Practical rules that match the CT section:
- eGFR <30 mL/min/1.73 m² or ongoing AKI: ask whether TOF MRA or noncontrast tests answer the question; if DSA is unavoidable (CTA-negative diffuse SAH), hydrate with isotonic saline and use the least contrast that still answers the anatomic question
- Iodinated allergy: premedication is a stopgap; MRA or a different modality is better when the indication is not immediately life-saving
- Gadolinium CE-MRA trades iodine for NSF risk in advanced CKD; TOF avoids both
- Metformin is held around iodine when eGFR is very low or AKI is present because of lactic-acidosis risk if contrast-associated injury occurs—not because metformin itself is nephrotoxic
Exam-style traps
- Declaring CTA-negative diffuse SAH a complete negative aneurysm workup
- Sending every unruptured 2 mm aneurysm on CTA straight to diagnostic DSA without a clinical indication
- Equating angiographic vasospasm on day 8 with a day-1 post-clip deficit
- Forgetting groin and stroke on the DSA consent while quoting only contrast nephropathy
- Using TOF MRA as proof that a 2 mm posterior communicating aneurysm is absent
- Stacking CTA plus CTP plus DSA in a patient with eGFR 18 without a kidney plan
Independent practice items at /practice/abim-neurocritical-care are a study bank for this reasoning, not the computer-based examination administered by ABPN.
Which modality is the gold-standard lumen study for characterizing a suspected brain AVM nidus, a dural AV fistula, or angiographic vasospasm that may be treated in the same sitting?
A 52-year-old woman has thunderclap headache. Noncontrast CT shows diffuse cisternal and sylvian SAH. CTA of the head does not show an aneurysm. What is the most appropriate vascular next step?
Before diagnostic cerebral DSA, which complication cluster should be discussed in addition to iodinated-contrast kidney injury?
On day 8 after coiling of a ruptured ACOM aneurysm, a patient develops a new aphasia. NCCT shows no hydrocephalus and no new blood. Which statement about vascular imaging is most accurate?