Role of Digital Subtraction Angiography (DSA) in Aneurysmal Subarachnoid Hemorrhage
DSA with 3-dimensional rotational angiography remains the gold standard for aneurysm detection and characterization in aSAH, and is mandatory when CTA is negative or inconclusive, when there is a diffuse hemorrhage pattern, or when detailed treatment planning is needed for endovascular therapy. 1, 2
Primary Indications for DSA
When CTA is Negative or Inconclusive:
- DSA must be performed when CTA fails to identify an aneurysm source in confirmed aSAH, particularly with diffuse hemorrhage patterns 1, 2
- In diffuse aSAH patterns with negative initial CTA, DSA detects causative lesions in approximately 13% of cases, often identifying atypical aneurysms (blister, dissecting, or fusiform types) 3
- Repeat delayed DSA is required if initial DSA is negative in diffuse patterns, detecting small aneurysms in 14% of these cases 1
For Small Aneurysm Detection:
- CTA unreliably demonstrates aneurysms <3 mm in size, making DSA essential when clinical suspicion remains high 1, 2
- 3D rotational angiography identifies aneurysms in 25% of patients with previously negative 2D angiograms 2
- DSA detected 15.6% of aneurysms missed by 2D imaging, with mean size of 2.79 mm 4
For Treatment Planning:
- DSA provides superior anatomic detail for determining whether an aneurysm is amenable to endovascular coiling versus surgical clipping 1, 2
- CTA may artificially widen the aneurysmal neck due to partial volume averaging, leading to erroneous conclusions about coiling feasibility 1
- 3D DSA is superior to CTA for detecting aneurysm neck characteristics, dome/neck ratios, and relationship to parent vessels and neighboring arteries 1, 4
Special Clinical Scenarios
Hemorrhage Pattern-Based Approach:
- Classic perimesencephalic SAH pattern with negative CTA may not require DSA (though this remains controversial) 1
- Diffuse aneurysmal pattern mandates DSA even after negative CTA 1, 3
- When blood is located in the sulci, DSA is recommended to confirm or exclude vasculitis 1
- Loss of consciousness accompanying hemorrhage warrants 2D and 3D cerebral angiography regardless of CTA findings 1
Detection of Secondary Pathology:
- DSA may reveal additional lesions not apparent on CTA, such as dural arteriovenous fistulas or vasospasm, that alter management 5
- 3D DSA has 100% sensitivity and 84% specificity for detecting vasospasm, superior to CTA 4
Technical Superiority of DSA
Anatomic Limitations of CTA:
- Overlying bone is problematic with CTA, especially at the skull base 1
- Vessel tortuosity decreases CTA specificity and can lead to misinterpretation 6
- Quality depends heavily on radiologist experience and scanner specifications (16- vs 64-detector rows) 1
DSA Advantages:
- 3D rotational angiography combined with 2D DSA provides the highest spatial resolution for morphological depiction of aneurysm anatomy 1
- DSA is always used in preparation for endovascular therapy 1
- Superior for detecting aneurysms in unusual locations (intracavernous carotid, posterior inferior cerebellar artery) 7
Clinical Algorithm
Step 1: Non-contrast CT confirms SAH 2
Step 2: High-quality CTA as initial vascular imaging 2
Step 3: Proceed to DSA if:
- CTA is negative with diffuse hemorrhage pattern 1, 3
- CTA shows aneurysm but endovascular treatment is being considered 1
- Hemorrhage accompanied by loss of consciousness 1
- Blood located in sulci (to exclude vasculitis) 1
- Aneurysm <3 mm suspected 1
- Initial DSA negative in diffuse pattern (repeat at 2 weeks) 1, 3
Step 4: Use 3D rotational angiography in addition to 2D DSA for complete evaluation 1, 2, 4
Critical Pitfalls to Avoid
- Do not rely solely on CTA for endovascular treatment planning—DSA provides essential information about neck morphology and parent vessel relationships 1
- Do not accept negative CTA in diffuse SAH patterns without DSA follow-up—atypical aneurysms are frequently missed 3
- Do not skip repeat delayed DSA when initial DSA is negative in diffuse patterns—14% will show aneurysms on delayed imaging 1
- Do not assume CTA quality is uniform across institutions—scanner specifications and processing algorithms significantly affect spatial resolution 1